{"document":{"aggregate_severity":{"namespace":"https://access.redhat.com/security/updates/classification/","text":"Critical"},"category":"csaf_security_advisory","csaf_version":"2.0","distribution":{"text":"Copyright © Red Hat, Inc. All rights reserved.","tlp":{"label":"WHITE","url":"https://www.first.org/tlp/"}},"lang":"en","notes":[{"category":"summary","text":"Red Hat AI Inference Model Optimization Tools 3.5.3 (CUDA) is now available.","title":"Topic"},{"category":"general","text":"Red Hat AI Inference Model Optimization Tools","title":"Details"},{"category":"legal_disclaimer","text":"This content is licensed under the Creative Commons Attribution 4.0 International License (https://creativecommons.org/licenses/by/4.0/). If you distribute this content, or a modified version of it, you must provide attribution to Red Hat Inc. and provide a link to the original.","title":"Terms of Use"}],"publisher":{"category":"vendor","contact_details":"https://access.redhat.com/security/team/contact/","issuing_authority":"Red Hat Product Security is responsible for vulnerability handling across all Red Hat products and services.","name":"Red Hat Product Security","namespace":"https://www.redhat.com"},"references":[{"category":"self","summary":"https://access.redhat.com/errata/RHSA-2026:94715","url":"https://access.redhat.com/errata/RHSA-2026:94715"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-111063","url":"https://access.redhat.com/security/cve/CVE-2026-111063"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-111064","url":"https://access.redhat.com/security/cve/CVE-2026-111064"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-111065","url":"https://access.redhat.com/security/cve/CVE-2026-111065"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-111066","url":"https://access.redhat.com/security/cve/CVE-2026-111066"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-111748","url":"https://access.redhat.com/security/cve/CVE-2026-111748"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-111750","url":"https://access.redhat.com/security/cve/CVE-2026-111750"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-112390","url":"https://access.redhat.com/security/cve/CVE-2026-112390"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-112391","url":"https://access.redhat.com/security/cve/CVE-2026-112391"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-112392","url":"https://access.redhat.com/security/cve/CVE-2026-112392"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-112393","url":"https://access.redhat.com/security/cve/CVE-2026-112393"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-112394","url":"https://access.redhat.com/security/cve/CVE-2026-112394"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-112395","url":"https://access.redhat.com/security/cve/CVE-2026-112395"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-112396","url":"https://access.redhat.com/security/cve/CVE-2026-112396"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-112397","url":"https://access.redhat.com/security/cve/CVE-2026-112397"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-112398","url":"https://access.redhat.com/security/cve/CVE-2026-112398"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-112399","url":"https://access.redhat.com/security/cve/CVE-2026-112399"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-112400","url":"https://access.redhat.com/security/cve/CVE-2026-112400"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-112401","url":"https://access.redhat.com/security/cve/CVE-2026-112401"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-112402","url":"https://access.redhat.com/security/cve/CVE-2026-112402"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-112403","url":"https://access.redhat.com/security/cve/CVE-2026-112403"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-112404","url":"https://access.redhat.com/security/cve/CVE-2026-112404"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-112405","url":"https://access.redhat.com/security/cve/CVE-2026-112405"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-112406","url":"https://access.redhat.com/security/cve/CVE-2026-112406"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-112891","url":"https://access.redhat.com/security/cve/CVE-2026-112891"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-112892","url":"https://access.redhat.com/security/cve/CVE-2026-112892"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-112893","url":"https://access.redhat.com/security/cve/CVE-2026-112893"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-112894","url":"https://access.redhat.com/security/cve/CVE-2026-112894"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-112895","url":"https://access.redhat.com/security/cve/CVE-2026-112895"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-112896","url":"https://access.redhat.com/security/cve/CVE-2026-112896"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-112897","url":"https://access.redhat.com/security/cve/CVE-2026-112897"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-112898","url":"https://access.redhat.com/security/cve/CVE-2026-112898"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-112899","url":"https://access.redhat.com/security/cve/CVE-2026-112899"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-112900","url":"https://access.redhat.com/security/cve/CVE-2026-112900"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-112901","url":"https://access.redhat.com/security/cve/CVE-2026-112901"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-112902","url":"https://access.redhat.com/security/cve/CVE-2026-112902"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-112940","url":"https://access.redhat.com/security/cve/CVE-2026-112940"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-114406","url":"https://access.redhat.com/security/cve/CVE-2026-114406"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-114501","url":"https://access.redhat.com/security/cve/CVE-2026-114501"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-96636","url":"https://access.redhat.com/security/cve/CVE-2026-96636"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-96637","url":"https://access.redhat.com/security/cve/CVE-2026-96637"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-96638","url":"https://access.redhat.com/security/cve/CVE-2026-96638"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-96639","url":"https://access.redhat.com/security/cve/CVE-2026-96639"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-96640","url":"https://access.redhat.com/security/cve/CVE-2026-96640"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-96641","url":"https://access.redhat.com/security/cve/CVE-2026-96641"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-96642","url":"https://access.redhat.com/security/cve/CVE-2026-96642"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-96643","url":"https://access.redhat.com/security/cve/CVE-2026-96643"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-96644","url":"https://access.redhat.com/security/cve/CVE-2026-96644"},{"category":"external","summary":"https://access.redhat.com/security/cve/CVE-2026-96778","url":"https://access.redhat.com/security/cve/CVE-2026-96778"},{"category":"external","summary":"https://access.redhat.com/security/updates/classification/","url":"https://access.redhat.com/security/updates/classification/"},{"category":"external","summary":"https://www.redhat.com/en/products/ai/inference-server","url":"https://www.redhat.com/en/products/ai/inference-server"},{"category":"self","summary":"Canonical URL","url":"https://security.access.redhat.com/data/csaf/v2/advisories/2026/rhsa-2026_94715.json"}],"title":"Red Hat Security Advisory: Red Hat AI Inference Model Optimization Tools 3.5.3 (CUDA)","tracking":{"current_release_date":"2026-11-04T14:06:47+00:00","generator":{"date":"2026-11-04T14:06:47+00:00","engine":{"name":"Red Hat SDEngine","version":"5.4.4"}},"id":"RHSA-2026:94715","initial_release_date":"2026-11-03T19:52:15+00:00","revision_history":[{"date":"2026-11-03T19:52:15+00:00","number":"1","summary":"Initial version"},{"date":"2026-11-03T19:52:26+00:00","number":"2","summary":"Last updated version"},{"date":"2026-11-04T14:06:47+00:00","number":"3","summary":"Last generated version"}],"status":"final","version":"3"}},"product_tree":{"branches":[{"branches":[{"branches":[{"category":"product_name","name":"Red Hat AI Inference Server 3.5","product":{"name":"Red Hat AI Inference Server 3.5","product_id":"Red Hat AI Inference Server 3.5","product_identification_helper":{"cpe":"cpe:/a:redhat:ai_inference_server:3.5::el9"}}}],"category":"product_family","name":"Red Hat AI Inference Server"},{"branches":[{"category":"product_version","name":"registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","product":{"name":"registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","product_id":"registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","product_identification_helper":{"purl":"pkg:oci/model-opt-cuda-rhel9@sha256%3A2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776?arch=amd64&repository_url=registry.redhat.io/rhaii/model-opt-cuda-rhel9&tag=1793444339"}}}],"category":"architecture","name":"amd64"},{"branches":[{"category":"product_version","name":"registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64","product":{"name":"registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64","product_id":"registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64","product_identification_helper":{"purl":"pkg:oci/model-opt-cuda-rhel9@sha256%3Ab81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16?arch=arm64&repository_url=registry.redhat.io/rhaii/model-opt-cuda-rhel9&tag=1793444339"}}}],"category":"architecture","name":"arm64"}],"category":"vendor","name":"Red Hat"}],"relationships":[{"category":"default_component_of","full_product_name":{"name":"registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64 as a component of Red Hat AI Inference Server 3.5","product_id":"Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64"},"product_reference":"registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","relates_to_product_reference":"Red Hat AI Inference Server 3.5"},{"category":"default_component_of","full_product_name":{"name":"registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64 as a component of Red Hat AI Inference Server 3.5","product_id":"Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"},"product_reference":"registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64","relates_to_product_reference":"Red Hat AI Inference Server 3.5"}]},"vulnerabilities":[{"cve":"CVE-2026-96636","cwe":{"id":"CWE-409","name":"Improper Handling of Highly Compressed Data (Data Amplification)"},"discovery_date":"2026-10-21T21:29:02.337984+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2558578"}],"notes":[{"category":"description","text":"A flaw was found in Pillow. A remote attacker could exploit this vulnerability by enticing an application or user to process a specially crafted multi-frame image file. Because decompression bomb protections are bypassed when navigating subsequent image frames, decoding the file allocates excessive system memory, resulting in a Denial of Service (DoS).","title":"Vulnerability description"},{"category":"summary","text":"pillow: pillow: Denial of Service via decompression bomb check bypass in multi-frame images","title":"Vulnerability summary"},{"category":"other","text":"Red Hat has evaluated this vulnerability as having Moderate security impact because successful exploitation leads strictly to denial of service via memory exhaustion, without enabling data corruption or remote code execution. In typical Red Hat deployments containing Python imaging components, applications are exposed only when explicitly configured to ingest and navigate through multi-frame images supplied by untrusted external sources. Because default system operations do not automatically iterate over secondary frames in such formats, the risk remains restricted to specific image-processing workloads.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-96636"},{"category":"external","summary":"RHBZ#2558578","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2558578"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-96636","url":"https://www.cve.org/CVERecord?id=CVE-2026-96636"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-96636","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-96636"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/commit/f1acfc54c53502948d0d14121872061df931db48","url":"https://github.com/python-pillow/Pillow/commit/f1acfc54c53502948d0d14121872061df931db48"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/pull/10011","url":"https://github.com/python-pillow/Pillow/pull/10011"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0","url":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-p8hx-v2p6-rp84","url":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-p8hx-v2p6-rp84"}],"release_date":"2026-10-21T21:05:20.850000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"}],"scores":[{"cvss_v3":{"attackComplexity":"LOW","attackVector":"NETWORK","availabilityImpact":"HIGH","baseScore":6.5,"baseSeverity":"MEDIUM","confidentialityImpact":"NONE","integrityImpact":"NONE","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"REQUIRED","vectorString":"CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:N/I:N/A:H","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Moderate"}],"title":"pillow: pillow: Denial of Service via decompression bomb check bypass in multi-frame images"},{"cve":"CVE-2026-96637","cwe":{"id":"CWE-409","name":"Improper Handling of Highly Compressed Data (Data Amplification)"},"discovery_date":"2026-10-21T21:23:17.500415+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2558574"}],"notes":[{"category":"description","text":"A flaw was found in Pillow. When processing multi-frame SPIDER images, decompression bomb safeguards are not reapplied when navigating to subsequent frames that declare excessively large dimensions. A remote attacker could exploit this vulnerability by providing a specially crafted image file, leading to excessive memory consumption and a Denial of Service (DoS).","title":"Vulnerability description"},{"category":"summary","text":"pillow: pillow: Denial of Service via decompression bomb bypass in SPIDER images","title":"Vulnerability summary"},{"category":"other","text":"Red Hat has evaluated this vulnerability as having a Moderate severity impact because exploitation requires an application to parse multi-frame SPIDER image files and advance through secondary frames, limiting the consequence to memory exhaustion and denial of service without compromising confidentiality or integrity. The SPIDER image format is specialized and not typically accepted or processed by general enterprise workloads or default service configurations. Consequently, typical deployments remain unaffected unless custom application workflows explicitly ingest and iterate through untrusted multi-frame SPIDER datasets.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-96637"},{"category":"external","summary":"RHBZ#2558574","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2558574"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-96637","url":"https://www.cve.org/CVERecord?id=CVE-2026-96637"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-96637","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-96637"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/commit/7cb0e02389d8aac2a60896a490f572db0b07b841","url":"https://github.com/python-pillow/Pillow/commit/7cb0e02389d8aac2a60896a490f572db0b07b841"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/pull/10076","url":"https://github.com/python-pillow/Pillow/pull/10076"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0","url":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-g842-9rc4-69pw","url":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-g842-9rc4-69pw"}],"release_date":"2026-10-21T21:01:20.249000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"}],"scores":[{"cvss_v3":{"attackComplexity":"LOW","attackVector":"NETWORK","availabilityImpact":"HIGH","baseScore":6.5,"baseSeverity":"MEDIUM","confidentialityImpact":"NONE","integrityImpact":"NONE","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"REQUIRED","vectorString":"CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:N/I:N/A:H","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Moderate"}],"title":"pillow: pillow: Denial of Service via decompression bomb bypass in SPIDER images"},{"cve":"CVE-2026-96638","cwe":{"id":"CWE-409","name":"Improper Handling of Highly Compressed Data (Data Amplification)"},"discovery_date":"2026-10-21T21:06:55.104329+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2558568"}],"notes":[{"category":"description","text":"A flaw was found in Pillow. When processing run-length encoded (RLE) bitmap images, the decoder fails to properly limit memory allocation for certain escape sequences and data runs. A remote attacker could exploit this flaw by supplying a specially crafted BMP image file to an application, causing excessive memory consumption and resulting in a Denial of Service (DoS).","title":"Vulnerability description"},{"category":"summary","text":"pillow: pillow: Denial of Service via crafted RLE BMP image","title":"Vulnerability summary"},{"category":"other","text":"Red Hat rates this vulnerability as Moderate impact because exploitation is constrained to a denial of service via memory exhaustion, with no compromise to system confidentiality or integrity. Within Red Hat environments, Pillow is included as a supporting library across containerized platforms and automation tooling, where triggering the flaw requires an application to ingest and decode untrusted bitmap files. Systems and services that do not process untrusted image uploads or run-length encoded graphics are not exposed to this risk.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-96638"},{"category":"external","summary":"RHBZ#2558568","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2558568"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-96638","url":"https://www.cve.org/CVERecord?id=CVE-2026-96638"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-96638","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-96638"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/commit/e7794835e323ff2691d9b3a974cade1a44301b9a","url":"https://github.com/python-pillow/Pillow/commit/e7794835e323ff2691d9b3a974cade1a44301b9a"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/pull/10015","url":"https://github.com/python-pillow/Pillow/pull/10015"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0","url":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-24rc-8hj2-v98c","url":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-24rc-8hj2-v98c"}],"release_date":"2026-10-21T20:47:21.767000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"}],"scores":[{"cvss_v3":{"attackComplexity":"LOW","attackVector":"NETWORK","availabilityImpact":"HIGH","baseScore":6.5,"baseSeverity":"MEDIUM","confidentialityImpact":"NONE","integrityImpact":"NONE","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"REQUIRED","vectorString":"CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:N/I:N/A:H","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Moderate"}],"title":"pillow: pillow: Denial of Service via crafted RLE BMP image"},{"cve":"CVE-2026-96639","cwe":{"id":"CWE-787","name":"Out-of-bounds Write"},"discovery_date":"2026-10-21T21:24:08.478957+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2558575"}],"notes":[{"category":"description","text":"A flaw was found in Pillow. When processing multi-frame images that transition between different color modes, Pillow reuses an undersized native memory buffer without verifying that its format matches the subsequent frame. A remote attacker can exploit this flaw by providing a specially crafted multi-frame image file, causing an out-of-bounds memory write. Successful exploitation can lead to arbitrary code execution or a Denial of Service (DoS) resulting from an application crash.","title":"Vulnerability description"},{"category":"summary","text":"pillow: Pillow multi-frame loaders can reuse stale native image storage after mode transitions","title":"Vulnerability summary"},{"category":"other","text":"This vulnerability is rated Critical because network-facing services that ingest and process untrusted multi-frame images can be exploited without privileges or user interaction, potentially resulting in memory corruption and remote code execution. Applications parsing sequential frames in formats such as MPO, PSD, or DCX are exposed when handling transitions between varying color modes within the same file. Red Hat Enterprise Linux releases distributing versions prior to Pillow 11.0.0 are unaffected as they lack the affected storage reuse implementation.","title":"Statement"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-96639"},{"category":"external","summary":"RHBZ#2558575","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2558575"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-96639","url":"https://www.cve.org/CVERecord?id=CVE-2026-96639"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-96639","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-96639"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/commit/78d97d994579aef81aafe123773d525f40f2522a","url":"https://github.com/python-pillow/Pillow/commit/78d97d994579aef81aafe123773d525f40f2522a"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/pull/10105","url":"https://github.com/python-pillow/Pillow/pull/10105"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0","url":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-v9pv-5gc8-w2m6","url":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-v9pv-5gc8-w2m6"}],"release_date":"2026-10-21T21:06:42.111000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"}],"threats":[{"category":"impact","details":"Important"}],"title":"pillow: Pillow multi-frame loaders can reuse stale native image storage after mode transitions"},{"cve":"CVE-2026-96640","cwe":{"id":"CWE-770","name":"Allocation of Resources Without Limits or Throttling"},"discovery_date":"2026-10-21T21:25:20.244307+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2558576"}],"notes":[{"category":"description","text":"A flaw was found in Pillow. When appending images to an existing PDF document, the PDF parser does not track duplicate indirect references within the page structure. A remote attacker can provide a specially crafted PDF file containing repeated references to the same page nodes, causing the application to process an exponential number of pages. This can consume excessive CPU resources, resulting in a Denial of Service (DoS).","title":"Vulnerability description"},{"category":"summary","text":"pillow: pillow: Denial of Service via duplicate page references in PDF files","title":"Vulnerability summary"},{"category":"other","text":"Red Hat has evaluated this vulnerability as having an Important severity rating because remote, unauthenticated attackers can cause severe application denial of service through exponential CPU exhaustion. Although Pillow is widely distributed across Red Hat products and container images, exploitation is constrained to environments that accept untrusted PDF files and append image content to them. Standard image processing pipelines and read operations remain unaffected by this issue.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-96640"},{"category":"external","summary":"RHBZ#2558576","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2558576"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-96640","url":"https://www.cve.org/CVERecord?id=CVE-2026-96640"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-96640","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-96640"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/commit/6f0e7ee1f9c2fe227906f8a4b926e6622b702bc7","url":"https://github.com/python-pillow/Pillow/commit/6f0e7ee1f9c2fe227906f8a4b926e6622b702bc7"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/pull/10093","url":"https://github.com/python-pillow/Pillow/pull/10093"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0","url":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-mrfp-p49p-3hf2","url":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-mrfp-p49p-3hf2"}],"release_date":"2026-10-21T21:03:22.785000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"}],"scores":[{"cvss_v3":{"attackComplexity":"LOW","attackVector":"NETWORK","availabilityImpact":"HIGH","baseScore":7.5,"baseSeverity":"HIGH","confidentialityImpact":"NONE","integrityImpact":"NONE","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"NONE","vectorString":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Important"}],"title":"pillow: pillow: Denial of Service via duplicate page references in PDF files"},{"cve":"CVE-2026-96641","cwe":{"id":"CWE-190","name":"Integer Overflow or Wraparound"},"discovery_date":"2026-10-21T21:18:01.846902+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2558573"}],"notes":[{"category":"description","text":"A flaw was found in Pillow. When processing text dimensions with certain fonts, negative character spacing values are not properly validated during text-width calculation, causing an integer overflow. A remote attacker could exploit this vulnerability by providing crafted text or font data to an application that calculates text layout or dimensions. Successful exploitation could lead to excessive resource allocation, resulting in a Denial of Service (DoS) or application logic errors.","title":"Vulnerability description"},{"category":"summary","text":"pillow: pillow: Denial of Service via integer overflow in text-width calculation","title":"Vulnerability summary"},{"category":"other","text":"This vulnerability is rated as Moderate impact because successful exploitation typically results in application denial of service or computational discrepancies rather than remote code execution. In typical Red Hat deployments across platforms such as Red Hat Enterprise Linux and OpenShift AI, Pillow is used by backend services and automated pipelines, which rarely expose raw font parsing or text-width metrics directly to unauthenticated external inputs. An attacker must supply specially crafted font files or text strings to an application performing dimension metrics to trigger the integer overflow.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-96641"},{"category":"external","summary":"RHBZ#2558573","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2558573"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-96641","url":"https://www.cve.org/CVERecord?id=CVE-2026-96641"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-96641","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-96641"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/commit/80de4be9e38d3a6148dd4d8491eb80cd12b760a4","url":"https://github.com/python-pillow/Pillow/commit/80de4be9e38d3a6148dd4d8491eb80cd12b760a4"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/pull/10010","url":"https://github.com/python-pillow/Pillow/pull/10010"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0","url":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-4rm8-3hmm-39fg","url":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-4rm8-3hmm-39fg"}],"release_date":"2026-10-21T20:51:24.388000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"}],"scores":[{"cvss_v3":{"attackComplexity":"LOW","attackVector":"NETWORK","availabilityImpact":"HIGH","baseScore":6.5,"baseSeverity":"MEDIUM","confidentialityImpact":"NONE","integrityImpact":"NONE","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"REQUIRED","vectorString":"CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:N/I:N/A:H","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Moderate"}],"title":"pillow: pillow: Denial of Service via integer overflow in text-width calculation"},{"cve":"CVE-2026-96642","cwe":{"id":"CWE-835","name":"Loop with Unreachable Exit Condition"},"discovery_date":"2026-10-21T21:12:46.782670+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2558571"}],"notes":[{"category":"description","text":"A flaw was found in Pillow. When an application processes tar archives using the TarIO component, an archive containing an invalid negative member size causes the parser to enter an infinite loop by repeatedly reading the same header. An attacker could exploit this by providing a specially crafted archive, leading to high CPU consumption and a Denial of Service (DoS).","title":"Vulnerability description"},{"category":"summary","text":"pillow: Pillow: Denial of Service via negative archive member size in TarIO","title":"Vulnerability summary"},{"category":"other","text":"This vulnerability is rated as Moderate impact because, while processing a crafted archive can trigger an infinite loop and complete CPU exhaustion, exploitation requires an application to explicitly invoke the TarIO helper class rather than standard image loading interfaces. Typical deployment workflows and default image parsing pathways do not utilize TarIO, significantly restricting exposure across Red Hat environments. The risk of denial of service is therefore confined to applications that directly consume untrusted archives via this specialized component.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-96642"},{"category":"external","summary":"RHBZ#2558571","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2558571"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-96642","url":"https://www.cve.org/CVERecord?id=CVE-2026-96642"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-96642","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-96642"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/commit/19b695fdcc5eddc37374e2b1645fc4185edaf71d","url":"https://github.com/python-pillow/Pillow/commit/19b695fdcc5eddc37374e2b1645fc4185edaf71d"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/pull/10012","url":"https://github.com/python-pillow/Pillow/pull/10012"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0","url":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-3j72-p6qg-6xp7","url":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-3j72-p6qg-6xp7"}],"release_date":"2026-10-21T20:49:56.989000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"}],"scores":[{"cvss_v3":{"attackComplexity":"LOW","attackVector":"NETWORK","availabilityImpact":"HIGH","baseScore":6.5,"baseSeverity":"MEDIUM","confidentialityImpact":"NONE","integrityImpact":"NONE","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"REQUIRED","vectorString":"CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:N/I:N/A:H","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Moderate"}],"title":"pillow: Pillow: Denial of Service via negative archive member size in TarIO"},{"cve":"CVE-2026-96643","cwe":{"id":"CWE-606","name":"Unchecked Input for Loop Condition"},"discovery_date":"2026-10-21T21:03:21.143962+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2558566"}],"notes":[{"category":"description","text":"A flaw was found in Pillow. When processing a specially crafted image with repeated Exchangeable Image File Format (EXIF) metadata headers, the library repeatedly copies and parses the data. A remote attacker could exploit this vulnerability by providing a malicious image file to an application that processes image metadata, causing excessive CPU consumption that results in a Denial of Service (DoS).","title":"Vulnerability description"},{"category":"summary","text":"pillow: pillow: Denial of Service via repeated EXIF headers","title":"Vulnerability summary"},{"category":"other","text":"Red Hat rates this vulnerability as having an Important severity because remote, unauthenticated attackers can cause severe denial of service by submitting crafted images to automated image-processing pipelines. Services utilizing Pillow to inspect metadata or perform operations such as orientation adjustments can experience quadratic CPU consumption when parsing repetitive header markers. Because this condition can be triggered remotely without privileges in standard image handling workflows, it poses an elevated availability risk to backend services.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-96643"},{"category":"external","summary":"RHBZ#2558566","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2558566"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-96643","url":"https://www.cve.org/CVERecord?id=CVE-2026-96643"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-96643","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-96643"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/commit/806e38381c781f32bbd22f4cd026678acd40fd6c","url":"https://github.com/python-pillow/Pillow/commit/806e38381c781f32bbd22f4cd026678acd40fd6c"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/pull/10073","url":"https://github.com/python-pillow/Pillow/pull/10073"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0","url":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-g4qv-2jxx-769h","url":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-g4qv-2jxx-769h"}],"release_date":"2026-10-21T20:56:26.344000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"},{"category":"workaround","details":"Mitigation for this issue is either not available or the currently available options do not meet the Red Hat Product Security criteria comprising ease of use and deployment, applicability to widespread installation base, or stability.","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"scores":[{"cvss_v3":{"attackComplexity":"LOW","attackVector":"NETWORK","availabilityImpact":"HIGH","baseScore":7.5,"baseSeverity":"HIGH","confidentialityImpact":"NONE","integrityImpact":"NONE","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"NONE","vectorString":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Important"}],"title":"pillow: pillow: Denial of Service via repeated EXIF headers"},{"cve":"CVE-2026-96644","cwe":{"id":"CWE-125","name":"Out-of-bounds Read"},"discovery_date":"2026-10-21T21:10:27.677012+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2558570"}],"notes":[{"category":"description","text":"A flaw was found in Pillow. When processing Windows Metafile (WMF) or Enhanced Metafile (EMF) images on Windows systems, the image parser improperly handles dimension values, allocating an undersized buffer while reading full dimensions. A remote attacker could exploit this vulnerability by enticing an application or user to process a specially crafted metafile, triggering an out-of-bounds memory read that results in a Denial of Service (DoS) via an application crash or potential information disclosure.","title":"Vulnerability description"},{"category":"summary","text":"pillow: Pillow: Denial of Service via out-of-bounds read in metafile parser","title":"Vulnerability summary"},{"category":"other","text":"Red Hat rates this vulnerability as an Important severity flaw because an unauthenticated remote attacker could cause an application crash or expose heap memory contents by supplying a crafted image file. In Red Hat Enterprise Linux environments, however, the vulnerable native metafile rendering code is compiled out, as the handler is restricted to Windows operating systems. Consequently, standard Red Hat enterprise installations are not susceptible to this memory over-read condition.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-96644"},{"category":"external","summary":"RHBZ#2558570","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2558570"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-96644","url":"https://www.cve.org/CVERecord?id=CVE-2026-96644"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-96644","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-96644"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/commit/4cf46bee338304c5c8968fd7f132ba0d40c428b7","url":"https://github.com/python-pillow/Pillow/commit/4cf46bee338304c5c8968fd7f132ba0d40c428b7"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/pull/10120","url":"https://github.com/python-pillow/Pillow/pull/10120"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0","url":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-9v9v-9qvf-h2hj","url":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-9v9v-9qvf-h2hj"}],"release_date":"2026-10-21T20:54:43.537000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"},{"category":"workaround","details":"Mitigation for this issue is either not available or the currently available options do not meet the Red Hat Product Security criteria comprising ease of use and deployment, applicability to widespread installation base, or stability.","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"scores":[{"cvss_v3":{"attackComplexity":"LOW","attackVector":"NETWORK","availabilityImpact":"HIGH","baseScore":7.1,"baseSeverity":"HIGH","confidentialityImpact":"LOW","integrityImpact":"NONE","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"REQUIRED","vectorString":"CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:L/I:N/A:H","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Important"}],"title":"pillow: Pillow: Denial of Service via out-of-bounds read in metafile parser"},{"cve":"CVE-2026-96778","cwe":{"id":"CWE-787","name":"Out-of-bounds Write"},"discovery_date":"2026-10-21T21:06:07.880492+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2558567"}],"notes":[{"category":"description","text":"A flaw was found in Pillow. A remote attacker could exploit a heap buffer overflow vulnerability to achieve arbitrary code execution or cause a Denial of Service (DoS). This issue occurs when an application re-encodes an untrusted image into JPEG format while preserving specially crafted color profile and comment metadata, causing an internal memory buffer to overflow. Successful exploitation allows adjacent memory to be overwritten, which can crash the application or permit code execution in suitable runtime environments.","title":"Vulnerability description"},{"category":"summary","text":"pillow: pillow: Arbitrary code execution via JPEG encoder buffer overflow","title":"Vulnerability summary"},{"category":"other","text":"Red Hat evaluates this vulnerability as having an Important severity because an unauthenticated attacker could trigger memory corruption leading to denial of service or arbitrary code execution if an exposed service processes untrusted graphics. In Red Hat ecosystems, Pillow is bundled across various containerized environments, including Red Hat OpenShift AI and Ansible Automation Platform, where background services or pipelines may ingest media files. However, exploitation requires an application workflow that explicitly re-encodes user-supplied images into JPEG format while retaining both embedded color profiles and comment metadata, which restricts the attack surface to specific image handling operations.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-96778"},{"category":"external","summary":"RHBZ#2558567","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2558567"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-96778","url":"https://www.cve.org/CVERecord?id=CVE-2026-96778"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-96778","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-96778"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/commit/02561b8dc6117d5670d4b469fecdd02c70854f9b","url":"https://github.com/python-pillow/Pillow/commit/02561b8dc6117d5670d4b469fecdd02c70854f9b"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/pull/10124","url":"https://github.com/python-pillow/Pillow/pull/10124"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0","url":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-g7cr-3p24-7wf7","url":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-g7cr-3p24-7wf7"}],"release_date":"2026-10-21T20:58:24.238000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"},{"category":"workaround","details":"Mitigation for this issue is either not available or the currently available options do not meet the Red Hat Product Security criteria comprising ease of use and deployment, applicability to widespread installation base, or stability.","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"scores":[{"cvss_v3":{"attackComplexity":"HIGH","attackVector":"NETWORK","availabilityImpact":"HIGH","baseScore":8.1,"baseSeverity":"HIGH","confidentialityImpact":"HIGH","integrityImpact":"HIGH","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"NONE","vectorString":"CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:H/I:H/A:H","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Important"}],"title":"pillow: pillow: Arbitrary code execution via JPEG encoder buffer overflow"},{"cve":"CVE-2026-111063","cwe":{"id":"CWE-319","name":"Cleartext Transmission of Sensitive Information"},"discovery_date":"2026-10-19T17:01:56.246752+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2555712"}],"notes":[{"category":"description","text":"A flaw was found in aiohttp. The application fails to properly enforce the Secure attribute for shared cookies. A remote attacker could exploit this vulnerability to cause cookies intended solely for encrypted connections to be transmitted in plaintext, resulting in sensitive information disclosure.","title":"Vulnerability description"},{"category":"summary","text":"aiohttp: aiohttp: Information disclosure via unenforced Secure attribute on shared cookies","title":"Vulnerability summary"},{"category":"other","text":"This vulnerability is rated as Low severity because successful exploitation requires an aiohttp client session to transmit requests across an unencrypted transport, alongside network interception capabilities to inspect the cleartext payload. In typical Red Hat environments, client sessions interacting with sensitive endpoints enforce TLS/HTTPS, restricting scenarios where secure cookies could be transmitted over cleartext. Consequently, the flaw presents only limited confidentiality risk and poses no threat to system integrity or availability.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-111063"},{"category":"external","summary":"RHBZ#2555712","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2555712"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-111063","url":"https://www.cve.org/CVERecord?id=CVE-2026-111063"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-111063","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-111063"},{"category":"external","summary":"https://github.com/aio-libs/aiohttp/commit/999c356d61e966faffa97b06ab0d0455226d36c5","url":"https://github.com/aio-libs/aiohttp/commit/999c356d61e966faffa97b06ab0d0455226d36c5"},{"category":"external","summary":"https://github.com/aio-libs/aiohttp/releases/tag/v3.14.4","url":"https://github.com/aio-libs/aiohttp/releases/tag/v3.14.4"},{"category":"external","summary":"https://github.com/aio-libs/aiohttp/security/advisories/GHSA-494j-54xm-rwp3","url":"https://github.com/aio-libs/aiohttp/security/advisories/GHSA-494j-54xm-rwp3"}],"release_date":"2026-10-19T16:44:31.331000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"},{"category":"workaround","details":"Mitigation for this issue is either not available or the currently available options do not meet the Red Hat Product Security criteria comprising ease of use and deployment, applicability to widespread installation base, or stability.","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"scores":[{"cvss_v3":{"attackComplexity":"HIGH","attackVector":"NETWORK","availabilityImpact":"NONE","baseScore":3.7,"baseSeverity":"LOW","confidentialityImpact":"LOW","integrityImpact":"NONE","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"NONE","vectorString":"CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:L/I:N/A:N","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Low"}],"title":"aiohttp: aiohttp: Information disclosure via unenforced Secure attribute on shared cookies"},{"cve":"CVE-2026-111064","cwe":{"id":"CWE-772","name":"Missing Release of Resource after Effective Lifetime"},"discovery_date":"2026-10-19T17:02:11.133402+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2555713"}],"notes":[{"category":"description","text":"A flaw was found in aiohttp. The server does not enforce a timeout during the initial phase of a request before any data is received. A remote attacker can exploit this vulnerability by keeping idle connections open indefinitely, exhausting available connection capacity and resulting in a Denial of Service (DoS).","title":"Vulnerability description"},{"category":"summary","text":"aiohttp: aiohttp: Denial of Service via missing initial request timeout","title":"Vulnerability summary"},{"category":"other","text":"This flaw affects products that run an aiohttp server exposed to untrusted clients. Products that use aiohttp only as an HTTP client are not affected.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-111064"},{"category":"external","summary":"RHBZ#2555713","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2555713"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-111064","url":"https://www.cve.org/CVERecord?id=CVE-2026-111064"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-111064","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-111064"},{"category":"external","summary":"https://github.com/aio-libs/aiohttp/commit/be846822b4b3a5a54b6163e794b42dc06fba54a3","url":"https://github.com/aio-libs/aiohttp/commit/be846822b4b3a5a54b6163e794b42dc06fba54a3"},{"category":"external","summary":"https://github.com/aio-libs/aiohttp/releases/tag/v3.14.4","url":"https://github.com/aio-libs/aiohttp/releases/tag/v3.14.4"},{"category":"external","summary":"https://github.com/aio-libs/aiohttp/security/advisories/GHSA-5wr8-gcqf-gqf3","url":"https://github.com/aio-libs/aiohttp/security/advisories/GHSA-5wr8-gcqf-gqf3"}],"release_date":"2026-10-19T16:47:35.036000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"},{"category":"workaround","details":"--- CME Mitigation Guidance (2026-10-20) ---\nDisclaimer: The following mitigation solutions are provided as a courtesy, pending the application of an official patch. This content was generated utilizing AI tools in conjunction with data from https://cmetaxonomy.org. Customers are advised to independently evaluate and test these temporary mitigations within their own infrastructure.\n\nThe highest-value action is to enforce initial keep-alive timeouts and first-request deadlines on newly established HTTP connections within aiohttp: upgrade to aiohttp 3.14.4 (or the fixed Ansible Automation Platform 24/Lightspeed Rhel/Rhoai/Odh Caikit Tgis Serving Rhel/Python3.11 Aiohttp/Koku/Exploit Intelligence/Vulnerability Analysis Rhel/Rhoai/Odh Trustyai Nemo Guardrails Server Rhel/Rhoai/Odh Kserve Storage Initializer Rhel/Rhoai/Odh Kserve Agent Rhel/Rhoai/Odh Kserve Controller Rhel/Rhoai/Odh Kserve Router Rhel/Rhaiis/Vllm Spyre Rhel/Rhaiis/Vllm Tpu Rhel/Cloudservices/Vulnerability Engine App/Rhoai/Odh Pipeline Runtime Datascience Cpu Py312 Rhel/Rhoai/Odh Pipeline Runtime Pytorch Cuda Py312 Rhel/Rhoai/Odh Pipeline Runtime Pytorch Rocm Py312 Rhel/Rhoai/Odh Pipeline Runtime Tensorflow Cuda Py312 Rhel/Rhoai/Odh Pipeline Runtime Tensorflow Rocm Py312 Rhel/Rhoai/Odh Workbench Jupyter Datascience Cpu Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Cuda Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Llmcompressor Cuda Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Rocm Py312 Rhel/Rhoai/Odh Workbench Jupyter Tensorflow Cuda Py312 Rhel/Rhoai/Odh Workbench Jupyter Tensorflow Rocm Py312 Rhel/Rhoai/Odh Workbench Jupyter Trustyai Cpu Py312 Rhel/Rhaiis/Vllm Cpu Rhel/Rhaiis/Vllm Neuron Rhel/Ansible Automation Platform 24/Controller Rhel/Ansible Automation Platform 25/Lightspeed Chatbot Rhel/Cloudservices/Insights Engine/Rhoai/Odh Ml Pipelines Runtime Generic Rhel/Rhoai/Odh Autorag Rhel/Rhoai/Odh Ta Lmes Job Rhel/Rhoai/Odh Th06 Cpu Torch291 Py312 Rhel/Rhoai/Odh Th06 Cuda130 Torch291 Py312 Rhel/Rhoai/Odh Training Cuda128 Torch29 Py312 Rhel/Rhoai/Odh Training Rocm64 Torch29 Py312 Rhel/Rhoai/Odh Vllm Gaudi Rhel/Rhai Early Access/Docling Serve Cuda Rhel/Automation Controller/Cloudservices/Floorist/Openshift Lightspeed/Lightspeed Ocp Rag Rhel/Rhoai/Odh Model Registry Job Async Upload Rhel/Rhaii/Vllm Gaudi Rhel/Rhaii/Vllm Neuron Rhel/Rhaii/Vllm Tpu Rhel/Rhoai/Odh Openvino Model Server Rhel/Rhoai/Odh Automl Rhel/Rhoai/Odh Kserve Autogluon Server Rhel/Rhoai/Odh Mlflow Rhel/Rhoai/Odh Llm D Kv Cache Rhel/Rhoai/Odh Th06 Cpu Torch210 Py312 Rhel/Rhoai/Odh Th06 Cuda130 Torch210 Py312 Rhel/Rhoai/Odh Th06 Rocm64 Torch291 Py312 Rhel/Ask Sre/Rhelai3/Bootc Gaudi Rhel/Ansible Automation Platform Tech Preview/Ee Minimal Rhel/Rhoai/Odh Training Cuda121 Torch24 Py311 Rhel/Rhoai/Odh Training Cuda124 Torch25 Py311 Rhel/Rhoai/Odh Training Cuda128 Torch28 Py312 Rhel/Rhoai/Odh Training Rocm62 Torch24 Py311 Rhel/Rhoai/Odh Training Rocm62 Torch25 Py311 Rhel/Rhoai/Odh Training Rocm64 Torch28 Py312 Rhel/Rhoai/Odh Trustyai Garak Lls Provider Dsp Rhel/Astro Virtual Assistant V/Lightspeed Core/Lightspeed Stack Rhel/Lightspeed Core/Rag Tool Cuda 12.9 Rhel/Mta/Mta Solution Server Rhel/Rhoai/Odh Feature Server Rhel/Rhoai/Odh Llama Stack Core Rhel/Rhoai/Odh Pipeline Runtime Pytorch Llmcompressor Cuda Py312 Rhel/Rhoai/Odh Mlserver Rhel/Rhai/Guidellm Rhel/Rhoai/Odh Pipeline Runtime Minimal Cpu Py312 Rhel/Rhoai/Odh Th Torch Cpu Py312 Rhel/Rhoai/Odh Th Torch Cuda Py312 Rhel/Rhoai/Odh Th Torch Rocm Py312 Rhel/Rhoai/Odh Workbench Jupyter Minimal Cpu Py312 Rhel/Rhoai/Odh Workbench Jupyter Minimal Cuda Py312 Rhel/Rhoai/Odh Workbench Jupyter Minimal Rocm Py312 Rhel/Rhaiis/Model Opt Cuda Rhel/Rhaiis/Vllm Cuda Rhel/Rhaiis/Vllm Rocm Rhel/Rhaii/Model Opt Cuda Rhel/Rhaii/Vllm Cpu Rhel/Rhaii/Vllm Cuda Rhel/Rhaii/Vllm Rocm Rhel/Rhaii/Vllm Spyre Rhel/Ansible Automation Platform 25/Controller Rhel/Ansible Automation Platform 25/Lightspeed Rhel/Ansible Automation Platform 26/Ansible Dev Tools Rhel/Ansible Automation Platform 26/Controller Rhel/Ansible Automation Platform 26/De Minimal Rhel/Ansible Automation Platform 26/De Supported Rhel/Ansible Automation Platform 26/Ee Minimal Rhel/Ansible Automation Platform 26/Ee Supported Rhel/Ansible Automation Platform 26/Hub Rhel/Ansible Automation Platform 26/Lightspeed Chatbot Rhel/Ansible Automation Platform 26/Lightspeed Rhel/Ansible Automation Platform 26/Platform Resource Runner Rhel/Ansible Automation Platform 27/Controller Rhel/Ansible Automation Platform 27/De Minimal Rhel/Ansible Automation Platform 27/De Supported Rhel/Ansible Automation Platform 27/Ee Control Plane Rhel/Ansible Automation Platform 27/Ee Minimal Rhel/Ansible Automation Platform 27/Ee Supported Rhel/Ansible Automation Platform 27/Hub Rhel/Ansible Automation Platform 27/Lightspeed Chatbot Rhel/Ansible Automation Platform 27/Lightspeed Rhel/Ansible Automation Platform 27/Platform Resource Runner Rhel/Ansible Automation Platform/Ee Minimal Rhel/Galaxy Ng/Automation Hub Container/Redhat User Workloads/Ansible Saas Ee/Ansible On Clouds/Gh Action Runner/Ansible Ai Connect Service/Insights Engine/Cloudservices/Subscription Central Automation/Google Lightspeed Agent/Lscore Deploy/Vulnerability Engine/Cloudservices/Sha Extractor/Lightspeed Core/Rag Tool Cpu Rhel/Openshift Lightspeed Tech Preview/Lightspeed Rag Tool Rhel/Openshift Lightspeed/Lightspeed Service Api Rhel/Rhoai/Odh Caikit Nlp Rhel/Rhoai/Odh Mlserver Cuda Rhel/Rhoai/Odh Workbench Codeserver Datascience Cpu Py312 Rhel/Discovery/Discovery Server Rhel/Rhelai3/Bootc Aws Cuda Rhel/Rhelai3/Bootc Azure Cuda Rhel/Rhelai3/Bootc Azure Rocm Rhel/Rhelai3/Bootc Cuda Rhel/Rhelai3/Bootc Gcp Cuda Rhel/Rhelai3/Bootc Rocm Rhel/Satellite/Iop Insights Engine Rhel/Satellite/Iop Vulnerability Engine Rhel/Python3.12 Aiohttp/Python Aiohttp/Insights Host Inventory/Aap Rag Content/Ansible Chatbot Stack/Aoc/Ee Aap Azure Sre/Ansible On Clouds/Aoc Utility/Redhat User Workloads/Release Service Utils Standalone/Lightspeed Services Ocp/Ocp Rules Rhel/App Sre/Managed Prometheus Exporter Initcontainer/App Sre/Managed Prometheus Exporter Base/Python3X Aiohttp builds), which schedule an initial keep-alive timeout check in connection_made() when keepalive_timeout is greater than zero, lower the default keepalive_timeout from 3630 to 75.0 seconds, and enforce a first-request deadline, closing all 2 injection vectors (idle TCP connections opened without sending request data, trickling partial request data before the first complete HTTP request). Until patched, this is a concern on aiohttp HTTP server deployments directly exposed to network traffic without intermediate timeout enforcement, so place aiohttp instances behind a reverse proxy that actively enforces connection read and idle timeouts before requests arrive. As a compensating measure, configure systemd WatchdogSec and automatic restart directives to monitor daemon responsiveness and restart the aiohttp service if connection exhaustion halts health checks. These controls limit blast radius and contain exploitation, but they do not fix the underlying missing initial request timeout defect — apply the upstream patch to fully remediate.\n\nHighest-impact CME: CME-906 Rate Limiting / Connection Throttling (high)","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"scores":[{"cvss_v3":{"attackComplexity":"LOW","attackVector":"NETWORK","availabilityImpact":"HIGH","baseScore":7.5,"baseSeverity":"HIGH","confidentialityImpact":"NONE","integrityImpact":"NONE","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"NONE","vectorString":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Important"}],"title":"aiohttp: aiohttp: Denial of Service via missing initial request timeout"},{"cve":"CVE-2026-111065","cwe":{"id":"CWE-772","name":"Missing Release of Resource after Effective Lifetime"},"discovery_date":"2026-10-19T17:52:19.540858+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2555741"}],"notes":[{"category":"description","text":"A flaw was found in aiohttp. An integer overflow in the processing of HTTP Max-Age directives can prevent connections from closing as intended. A remote attacker could exploit this vulnerability to exhaust the connection pool, leading to a temporary Denial of Service (DoS).","title":"Vulnerability description"},{"category":"summary","text":"aiohttp: aiohttp: Denial of Service via Max-Age processing overflow","title":"Vulnerability summary"},{"category":"other","text":"This vulnerability is rated as having a Moderate severity for Red Hat products because an attacker cannot compromise confidentiality or data integrity, and the resulting denial of service is limited to transient connector pool exhaustion that is recovered once garbage collection runs. Red Hat environments that rely on asynchronous HTTP clients, including components within Red Hat OpenShift AI and Red Hat Ansible Automation Platform, may experience temporary connection starvation when communicating with untrusted remote endpoints. Because exploitation does not cause persistent crashes or allow code execution, the overall blast radius remains constrained.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-111065"},{"category":"external","summary":"RHBZ#2555741","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2555741"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-111065","url":"https://www.cve.org/CVERecord?id=CVE-2026-111065"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-111065","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-111065"},{"category":"external","summary":"https://github.com/aio-libs/aiohttp/commit/9dd8eaf4e31f1237ef2c1574a8c82a739fca5805","url":"https://github.com/aio-libs/aiohttp/commit/9dd8eaf4e31f1237ef2c1574a8c82a739fca5805"},{"category":"external","summary":"https://github.com/aio-libs/aiohttp/releases/tag/v3.14.4","url":"https://github.com/aio-libs/aiohttp/releases/tag/v3.14.4"},{"category":"external","summary":"https://github.com/aio-libs/aiohttp/security/advisories/GHSA-6cjc-53gc-5rqm","url":"https://github.com/aio-libs/aiohttp/security/advisories/GHSA-6cjc-53gc-5rqm"}],"release_date":"2026-10-19T16:57:26.696000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"}],"scores":[{"cvss_v3":{"attackComplexity":"LOW","attackVector":"NETWORK","availabilityImpact":"LOW","baseScore":5.3,"baseSeverity":"MEDIUM","confidentialityImpact":"NONE","integrityImpact":"NONE","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"NONE","vectorString":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:L","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Moderate"}],"title":"aiohttp: aiohttp: Denial of Service via Max-Age processing overflow"},{"cve":"CVE-2026-111066","cwe":{"id":"CWE-770","name":"Allocation of Resources Without Limits or Throttling"},"discovery_date":"2026-10-19T17:23:14.934493+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2555720"}],"notes":[{"category":"description","text":"A flaw was found in aiohttp. When processing multipart requests, the framework fails to properly limit the number of open file descriptors retained during request handling. An unauthenticated remote attacker can exploit this issue by sending specially crafted multipart requests, exhausting available file descriptors and causing a Denial of Service (DoS).","title":"Vulnerability description"},{"category":"summary","text":"aiohttp: aiohttp: Denial of Service via file descriptor exhaustion in multipart request processing","title":"Vulnerability summary"},{"category":"other","text":"This vulnerability is rated as Important for Red Hat products because an unauthenticated remote attacker can cause a denial of service without requiring user interaction or elevated privileges. In Red Hat environments, exposure is limited to services or containerized applications built with the aiohttp framework that explicitly handle multipart form uploads. When such endpoints are targeted, excessive file handle consumption can exhaust operating system resources, preventing the affected service from handling legitimate requests.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-111066"},{"category":"external","summary":"RHBZ#2555720","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2555720"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-111066","url":"https://www.cve.org/CVERecord?id=CVE-2026-111066"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-111066","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-111066"},{"category":"external","summary":"https://github.com/aio-libs/aiohttp/commit/2d94633b7f1b4c1ba07d647b629a8440652b2d09","url":"https://github.com/aio-libs/aiohttp/commit/2d94633b7f1b4c1ba07d647b629a8440652b2d09"},{"category":"external","summary":"https://github.com/aio-libs/aiohttp/releases/tag/v3.14.4","url":"https://github.com/aio-libs/aiohttp/releases/tag/v3.14.4"},{"category":"external","summary":"https://github.com/aio-libs/aiohttp/security/advisories/GHSA-9xw5-w5jj-pvcg","url":"https://github.com/aio-libs/aiohttp/security/advisories/GHSA-9xw5-w5jj-pvcg"}],"release_date":"2026-10-19T16:59:47.217000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"},{"category":"workaround","details":"--- CME Mitigation Guidance (2026-10-20) ---\nDisclaimer: The following mitigation solutions are provided as a courtesy, pending the application of an official patch. This content was generated utilizing AI tools in conjunction with data from https://cmetaxonomy.org. Customers are advised to independently evaluate and test these temporary mitigations within their own infrastructure.\n\nThe highest-value action is to enforce strict limits on multipart payload sizes and uploaded part counts in request handlers invoking Request.post() to prevent unbounded file descriptor allocation: upgrade to aiohttp 3.14.4 (or the fixed Ansible Automation Platform 24/Lightspeed Rhel/Rhoai/Odh Caikit Tgis Serving Rhel/Python3.11 Aiohttp/Koku/Exploit Intelligence/Vulnerability Analysis Rhel/Rhoai/Odh Trustyai Nemo Guardrails Server Rhel/Rhoai/Odh Kserve Storage Initializer Rhel/Rhoai/Odh Kserve Agent Rhel/Rhoai/Odh Kserve Controller Rhel/Rhoai/Odh Kserve Router Rhel/Rhaiis/Vllm Spyre Rhel/Rhaiis/Vllm Tpu Rhel/Cloudservices/Vulnerability Engine App/Rhoai/Odh Pipeline Runtime Datascience Cpu Py312 Rhel/Rhoai/Odh Pipeline Runtime Pytorch Cuda Py312 Rhel/Rhoai/Odh Pipeline Runtime Pytorch Rocm Py312 Rhel/Rhoai/Odh Pipeline Runtime Tensorflow Cuda Py312 Rhel/Rhoai/Odh Pipeline Runtime Tensorflow Rocm Py312 Rhel/Rhoai/Odh Workbench Jupyter Datascience Cpu Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Cuda Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Llmcompressor Cuda Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Rocm Py312 Rhel/Rhoai/Odh Workbench Jupyter Tensorflow Cuda Py312 Rhel/Rhoai/Odh Workbench Jupyter Tensorflow Rocm Py312 Rhel/Rhoai/Odh Workbench Jupyter Trustyai Cpu Py312 Rhel/Rhaiis/Vllm Cpu Rhel/Rhaiis/Vllm Neuron Rhel/Ansible Automation Platform 24/Controller Rhel/Ansible Automation Platform 25/Lightspeed Chatbot Rhel/Cloudservices/Insights Engine/Rhoai/Odh Ml Pipelines Runtime Generic Rhel/Rhoai/Odh Autorag Rhel/Rhoai/Odh Ta Lmes Job Rhel/Rhoai/Odh Th06 Cpu Torch291 Py312 Rhel/Rhoai/Odh Th06 Cuda130 Torch291 Py312 Rhel/Rhoai/Odh Training Cuda128 Torch29 Py312 Rhel/Rhoai/Odh Training Rocm64 Torch29 Py312 Rhel/Rhoai/Odh Vllm Gaudi Rhel/Rhai Early Access/Docling Serve Cuda Rhel/Automation Controller/Cloudservices/Floorist/Openshift Lightspeed/Lightspeed Ocp Rag Rhel/Rhoai/Odh Model Registry Job Async Upload Rhel/Rhaii/Vllm Gaudi Rhel/Rhaii/Vllm Neuron Rhel/Rhaii/Vllm Tpu Rhel/Rhoai/Odh Openvino Model Server Rhel/Rhoai/Odh Automl Rhel/Rhoai/Odh Kserve Autogluon Server Rhel/Rhoai/Odh Mlflow Rhel/Rhoai/Odh Llm D Kv Cache Rhel/Rhoai/Odh Th06 Cpu Torch210 Py312 Rhel/Rhoai/Odh Th06 Cuda130 Torch210 Py312 Rhel/Rhoai/Odh Th06 Rocm64 Torch291 Py312 Rhel/Ask Sre/Rhelai3/Bootc Gaudi Rhel/Ansible Automation Platform Tech Preview/Ee Minimal Rhel/Rhoai/Odh Training Cuda121 Torch24 Py311 Rhel/Rhoai/Odh Training Cuda124 Torch25 Py311 Rhel/Rhoai/Odh Training Cuda128 Torch28 Py312 Rhel/Rhoai/Odh Training Rocm62 Torch24 Py311 Rhel/Rhoai/Odh Training Rocm62 Torch25 Py311 Rhel/Rhoai/Odh Training Rocm64 Torch28 Py312 Rhel/Rhoai/Odh Trustyai Garak Lls Provider Dsp Rhel/Astro Virtual Assistant V/Lightspeed Core/Lightspeed Stack Rhel/Lightspeed Core/Rag Tool Cuda 12.9 Rhel/Mta/Mta Solution Server Rhel/Rhoai/Odh Feature Server Rhel/Rhoai/Odh Llama Stack Core Rhel/Rhoai/Odh Pipeline Runtime Pytorch Llmcompressor Cuda Py312 Rhel/Rhoai/Odh Mlserver Rhel/Rhai/Guidellm Rhel/Rhoai/Odh Pipeline Runtime Minimal Cpu Py312 Rhel/Rhoai/Odh Th Torch Cpu Py312 Rhel/Rhoai/Odh Th Torch Cuda Py312 Rhel/Rhoai/Odh Th Torch Rocm Py312 Rhel/Rhoai/Odh Workbench Jupyter Minimal Cpu Py312 Rhel/Rhoai/Odh Workbench Jupyter Minimal Cuda Py312 Rhel/Rhoai/Odh Workbench Jupyter Minimal Rocm Py312 Rhel/Rhaiis/Model Opt Cuda Rhel/Rhaiis/Vllm Cuda Rhel/Rhaiis/Vllm Rocm Rhel/Rhaii/Model Opt Cuda Rhel/Rhaii/Vllm Cpu Rhel/Rhaii/Vllm Cuda Rhel/Rhaii/Vllm Rocm Rhel/Rhaii/Vllm Spyre Rhel/Ansible Automation Platform 25/Controller Rhel/Ansible Automation Platform 25/Lightspeed Rhel/Ansible Automation Platform 26/Ansible Dev Tools Rhel/Ansible Automation Platform 26/Controller Rhel/Ansible Automation Platform 26/De Minimal Rhel/Ansible Automation Platform 26/De Supported Rhel/Ansible Automation Platform 26/Ee Minimal Rhel/Ansible Automation Platform 26/Ee Supported Rhel/Ansible Automation Platform 26/Hub Rhel/Ansible Automation Platform 26/Lightspeed Chatbot Rhel/Ansible Automation Platform 26/Lightspeed Rhel/Ansible Automation Platform 26/Platform Resource Runner Rhel/Ansible Automation Platform 27/Controller Rhel/Ansible Automation Platform 27/De Minimal Rhel/Ansible Automation Platform 27/De Supported Rhel/Ansible Automation Platform 27/Ee Control Plane Rhel/Ansible Automation Platform 27/Ee Minimal Rhel/Ansible Automation Platform 27/Ee Supported Rhel/Ansible Automation Platform 27/Hub Rhel/Ansible Automation Platform 27/Lightspeed Chatbot Rhel/Ansible Automation Platform 27/Lightspeed Rhel/Ansible Automation Platform 27/Platform Resource Runner Rhel/Ansible Automation Platform/Ee Minimal Rhel/Galaxy Ng/Automation Hub Container/Redhat User Workloads/Ansible Saas Ee/Ansible On Clouds/Gh Action Runner/Ansible Ai Connect Service/Insights Engine/Cloudservices/Subscription Central Automation/Google Lightspeed Agent/Lscore Deploy/Vulnerability Engine/Cloudservices/Sha Extractor/Lightspeed Core/Rag Tool Cpu Rhel/Openshift Lightspeed Tech Preview/Lightspeed Rag Tool Rhel/Openshift Lightspeed/Lightspeed Service Api Rhel/Rhoai/Odh Caikit Nlp Rhel/Rhoai/Odh Mlserver Cuda Rhel/Rhoai/Odh Workbench Codeserver Datascience Cpu Py312 Rhel/Discovery/Discovery Server Rhel/Rhelai3/Bootc Aws Cuda Rhel/Rhelai3/Bootc Azure Cuda Rhel/Rhelai3/Bootc Azure Rocm Rhel/Rhelai3/Bootc Cuda Rhel/Rhelai3/Bootc Gcp Cuda Rhel/Rhelai3/Bootc Rocm Rhel/Satellite/Iop Insights Engine Rhel/Satellite/Iop Vulnerability Engine Rhel/Python3.12 Aiohttp/Python Aiohttp/Insights Host Inventory/Aap Rag Content/Ansible Chatbot Stack/Aoc/Ee Aap Azure Sre/Ansible On Clouds/Aoc Utility/Redhat User Workloads/Release Service Utils Standalone/Lightspeed Services Ocp/Ocp Rules Rhel/App Sre/Managed Prometheus Exporter Initcontainer/App Sre/Managed Prometheus Exporter Base/Python3X Aiohttp builds), which replace tempfile.TemporaryFile with SpooledTemporaryFile to buffer file parts in memory up to 1 MB and enforce client_max_size on total received payload bytes, closing multipart HTTP POST requests containing numerous small file parts. Until patched, this is a concern on Network-accessible aiohttp HTTP servers where request handlers invoke Request.post() without authentication, so Restrict network access to unauthenticated HTTP endpoints that invoke Request.post(), or place an upstream proxy to drop multipart POST requests with excessive part counts. As a compensating measure, Configure systemd WatchdogSec and process Restart directives to detect hung processes and automatically restart instances during file descriptor starvation and Apply rate limiting and connection throttling on incoming HTTP POST requests by client IP to impede rapid exhaustion attempts. These controls limit blast radius and contain exploitation, but they do not fix the underlying file descriptor exhaustion defect — apply the upstream patch to fully remediate.\n\nHighest-impact CME: CME-1311 Input Size and Quantity Bounds Enforcement (high)","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"scores":[{"cvss_v3":{"attackComplexity":"LOW","attackVector":"NETWORK","availabilityImpact":"HIGH","baseScore":7.5,"baseSeverity":"HIGH","confidentialityImpact":"NONE","integrityImpact":"NONE","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"NONE","vectorString":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Important"}],"title":"aiohttp: aiohttp: Denial of Service via file descriptor exhaustion in multipart request processing"},{"cve":"CVE-2026-111748","cwe":{"id":"CWE-348","name":"Use of Less Trusted Source"},"discovery_date":"2026-10-19T17:23:35.790047+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2555722"}],"notes":[{"category":"description","text":"A flaw was found in aiohttp. A remote attacker can bypass security controls by sending an absolute-form HTTP request that spoofs the connection protocol scheme. This issue occurs because the framework evaluates attacker-controlled request-target data rather than verifying the underlying connection protocol. Consequently, applications that rely on the reported scheme to make security decisions may mistakenly treat unencrypted connections as encrypted, potentially leading to unauthorized access or policy circumvention.","title":"Vulnerability description"},{"category":"summary","text":"aiohttp: aiohttp: Security control bypass via request scheme spoofing","title":"Vulnerability summary"},{"category":"other","text":"This vulnerability is rated as Moderate impact because successful exploitation requires an application's internal security logic to explicitly trust the request scheme attribute for policy decisions, which is uncommon across standard deployments. In Red Hat environments, such as Red Hat Ansible Automation Platform and Red Hat OpenShift AI, services typically operate behind ingress controllers, API gateways, or reverse proxies that independently enforce transport-layer security and TLS policies. Consequently, an attacker cannot bypass access boundaries unless custom application handlers directly depend on this spoofed request target attribute to determine whether a connection is secure.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-111748"},{"category":"external","summary":"RHBZ#2555722","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2555722"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-111748","url":"https://www.cve.org/CVERecord?id=CVE-2026-111748"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-111748","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-111748"},{"category":"external","summary":"https://github.com/aio-libs/aiohttp/commit/6d4e7b01f9b951aa084a75d7356e907d4ee80411","url":"https://github.com/aio-libs/aiohttp/commit/6d4e7b01f9b951aa084a75d7356e907d4ee80411"},{"category":"external","summary":"https://github.com/aio-libs/aiohttp/releases/tag/v3.14.4","url":"https://github.com/aio-libs/aiohttp/releases/tag/v3.14.4"},{"category":"external","summary":"https://github.com/aio-libs/aiohttp/security/advisories/GHSA-f9xw-jh8m-3mwq","url":"https://github.com/aio-libs/aiohttp/security/advisories/GHSA-f9xw-jh8m-3mwq"}],"release_date":"2026-10-19T17:02:21.289000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"}],"scores":[{"cvss_v3":{"attackComplexity":"LOW","attackVector":"NETWORK","availabilityImpact":"NONE","baseScore":6.5,"baseSeverity":"MEDIUM","confidentialityImpact":"LOW","integrityImpact":"LOW","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"NONE","vectorString":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:L/A:N","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Moderate"}],"title":"aiohttp: aiohttp: Security control bypass via request scheme spoofing"},{"cve":"CVE-2026-111750","cwe":{"id":"CWE-444","name":"Inconsistent Interpretation of HTTP Requests"},"discovery_date":"2026-10-19T17:22:56.880843+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2555719"}],"notes":[{"category":"description","text":"A flaw was found in aiohttp. When running with the pure Python parser instead of compiled extensions, the parser accepts ambiguous HTTP (Hypertext Transfer Protocol) request framing. An unauthenticated remote attacker could exploit this flaw by sending specially crafted requests through an intermediary proxy that interprets message boundaries differently, resulting in HTTP request smuggling. This may allow an attacker to bypass security controls, poison web caches, or manipulate requests from other users.","title":"Vulnerability description"},{"category":"summary","text":"aiohttp: aiohttp: HTTP request smuggling via ambiguous request framing","title":"Vulnerability summary"},{"category":"other","text":"This vulnerability is rated as Moderate severity because successful exploitation requires specific non-default configurations—specifically forcing the fallback pure-Python parser—alongside an upstream proxy that processes HTTP framing inconsistently. In standard Red Hat product deployments, aiohttp utilizes compiled native C extensions by default, which are not subject to this parsing flaw. Consequently, exploitation is restricted to environments where the C parser is explicitly disabled or unavailable, limiting the practical blast radius.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-111750"},{"category":"external","summary":"RHBZ#2555719","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2555719"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-111750","url":"https://www.cve.org/CVERecord?id=CVE-2026-111750"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-111750","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-111750"},{"category":"external","summary":"https://github.com/aio-libs/aiohttp/commit/9b27347e61454d9f8ca2055f4c67a37dee336e83","url":"https://github.com/aio-libs/aiohttp/commit/9b27347e61454d9f8ca2055f4c67a37dee336e83"},{"category":"external","summary":"https://github.com/aio-libs/aiohttp/releases/tag/v3.14.4","url":"https://github.com/aio-libs/aiohttp/releases/tag/v3.14.4"},{"category":"external","summary":"https://github.com/aio-libs/aiohttp/security/advisories/GHSA-pqxq-w44r-73xw","url":"https://github.com/aio-libs/aiohttp/security/advisories/GHSA-pqxq-w44r-73xw"}],"release_date":"2026-10-19T17:06:37.997000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"}],"scores":[{"cvss_v3":{"attackComplexity":"HIGH","attackVector":"NETWORK","availabilityImpact":"LOW","baseScore":4.8,"baseSeverity":"MEDIUM","confidentialityImpact":"NONE","integrityImpact":"LOW","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"NONE","vectorString":"CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:L/A:L","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Moderate"}],"title":"aiohttp: aiohttp: HTTP request smuggling via ambiguous request framing"},{"cve":"CVE-2026-112390","cwe":{"id":"CWE-835","name":"Loop with Unreachable Exit Condition"},"discovery_date":"2026-10-19T20:51:58.037704+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2555848"}],"notes":[{"category":"description","text":"A flaw was found in Pillow. This vulnerability allows a remote attacker to cause a Denial of Service (DoS) by supplying a specially crafted Microsoft Image Composer (MIC) file. When an application attempts to count or extract frames from the malicious image, an improper state reset triggers an infinite loop. This non-advancing loop consumes excessive CPU resources and can hang application worker processes indefinitely.","title":"Vulnerability description"},{"category":"summary","text":"pillow: Pillow: Denial of Service via infinite loop during crafted MIC image parsing","title":"Vulnerability summary"},{"category":"other","text":"This vulnerability is rated as an Important severity flaw because an unauthenticated remote attacker can induce an infinite loop, exhausting worker CPU resources and resulting in a denial of service for applications handling untrusted image data. Within Red Hat environments, Pillow is deployed across numerous container images, machine learning frameworks, and automated backend services. Exploitation requires the optional olefile dependency to be present to register the Microsoft Image Composer format handler, and specifically targets workflows that query frame counts or process multi-frame images. Environments that do not inspect frame counts or do not have olefile installed are not vulnerable.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-112390"},{"category":"external","summary":"RHBZ#2555848","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2555848"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-112390","url":"https://www.cve.org/CVERecord?id=CVE-2026-112390"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-112390","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-112390"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/commit/26cb3ed061173b817b9ee4b6f507cb0e1e293218","url":"https://github.com/python-pillow/Pillow/commit/26cb3ed061173b817b9ee4b6f507cb0e1e293218"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0","url":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-22hg-r758-rrw9","url":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-22hg-r758-rrw9"}],"release_date":"2026-10-19T20:35:39.189000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"},{"category":"workaround","details":"--- CME Mitigation Guidance (2026-10-20) ---\nDisclaimer: The following mitigation solutions are provided as a courtesy, pending the application of an official patch. This content was generated utilizing AI tools in conjunction with data from https://cmetaxonomy.org. Customers are advised to independently evaluate and test these temporary mitigations within their own infrastructure.\n\nThe highest-value action is to prevent Pillow from enumerating or extracting frames from untrusted Microsoft Image Composer (MIC) files when olefile is installed: upgrade to Pillow 13.0.0 (or the fixed Python Pillow/Python3.11 Pillow/Rhoai/Odh Openvino Model Server Rhel/Rhelai3/Bootc Cuda Rhel/Rhelai3/Bootc Gaudi Rhel/Rhelai3/Bootc Rocm Rhel/Rhelai3/Disk Image Cuda Rhel/Python3X Pillow/Rhaiis/Vllm Spyre Rhel/Rhoai/Odh Kserve Storage Initializer Rhel/Rhoai/Odh Kserve Agent Rhel/Rhoai/Odh Kserve Controller Rhel/Rhoai/Odh Kserve Router Rhel/Rhoai/Odh Feature Server Rhel/Rhoai/Odh Trustyai Nemo Guardrails Server Rhel/Rhaiis/Vllm Tpu Rhel/Rhoai/Odh Training Cuda128 Torch29 Py312 Rhel/Rhoai/Odh Training Rocm64 Torch29 Py312 Rhel/Rhaiis/Vllm Cpu Rhel/Rhaiis/Vllm Neuron Rhel/Ansible Automation Platform 25/Lightspeed Chatbot Rhel/Openshift Lightspeed/Lightspeed Ocp Rag Rhel/Rhoai/Odh Th06 Cpu Torch291 Py312 Rhel/Rhoai/Odh Th06 Cuda130 Torch291 Py312 Rhel/Rhoai/Odh Vllm Gaudi Rhel/Python3.12 Pillow/Rhaii/Vllm Gaudi Rhel/Rhaii/Vllm Neuron Rhel/Rhaii/Vllm Tpu Rhel/Lightspeed Core/Lightspeed Stack Rhel/Rhoai/Odh Kserve Autogluon Server Rhel/Rhai Early Access/Docling Sdk Cuda Rhel/Rhai Early Access/Docling Serve Cuda Rhel/Rhoai/Odh Llm D Kv Cache Rhel/Rhoai/Odh Th06 Cpu Torch210 Py312 Rhel/Rhoai/Odh Th06 Cuda130 Torch210 Py312 Rhel/Rhoai/Odh Th06 Rocm64 Torch291 Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Llmcompressor Cuda Py312 Rhel/Ask Sre/Rhel Cla/Rlsapi Rhel/Rhaiis/Model Opt Cuda Rhel/Rhaiis/Vllm Cuda Rhel/Rhaiis/Vllm Rocm Rhel/Rhaii/Model Opt Cuda Rhel/Rhaii/Vllm Cpu Rhel/Rhaii/Vllm Cuda Rhel/Rhaii/Vllm Rocm Rhel/Rhaii/Vllm Spyre Rhel/Ansible Automation Platform/Automation Dashboard Rhel/Ansible Automation Platform 26/Hub Rhel/Ansible Automation Platform 26/Lightspeed Chatbot Rhel/Ansible Automation Platform 27/Hub Rhel/Ansible Automation Platform 27/Lightspeed Chatbot Rhel/Galaxy Ng/Automation Hub Container/Aap Rag Content/Ansible Chatbot Stack/Koku Report Emailer/Exploit Intelligence/Vulnerability Analysis Rhel/Insights Host Inventory/Google Lightspeed Agent/Lscore Deploy/Lightspeed Core/Rag Tool Cpu Rhel/Lightspeed Core/Rag Tool Cuda 12.9 Rhel/Lightspeed Services Ocp/Ocp Rules Rhel/Openshift Lightspeed Tech Preview/Lightspeed Rag Tool Rhel/Openshift Lightspeed/Lightspeed Service Api Rhel/Rhoai/Odh Caikit Tgis Serving Rhel/Rhoai/Odh Llama Stack Core Rhel/Rhoai/Odh Pipeline Runtime Pytorch Llmcompressor Cuda Py312 Rhel/Rhoai/Odh Automl Rhel/Rhoai/Odh Autorag Rhel/Rhoai/Odh Mlflow Rhel/Rhoai/Odh Pipeline Runtime Datascience Cpu Py312 Rhel/Rhoai/Odh Pipeline Runtime Pytorch Cuda Py312 Rhel/Rhoai/Odh Pipeline Runtime Pytorch Rocm Py312 Rhel/Rhoai/Odh Pipeline Runtime Tensorflow Cuda Py312 Rhel/Rhoai/Odh Pipeline Runtime Tensorflow Rocm Py312 Rhel/Rhoai/Odh Trustyai Garak Lls Provider Dsp Rhel/Rhoai/Odh Workbench Codeserver Datascience Cpu Py312 Rhel/Rhoai/Odh Workbench Jupyter Datascience Cpu Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Cuda Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Rocm Py312 Rhel/Rhoai/Odh Workbench Jupyter Tensorflow Cuda Py312 Rhel/Rhoai/Odh Workbench Jupyter Tensorflow Rocm Py312 Rhel/Rhoai/Odh Workbench Jupyter Trustyai Cpu Py312 Rhel/Rhai/Guidellm Rhel/Rhoai/Odh Ogx Core Rhel/Rhoai/Odh Th Torch Cpu Py312 Rhel/Rhoai/Odh Th Torch Cuda Py312 Rhel/Rhoai/Odh Th Torch Rocm Py312 Rhel/Quay/Quay Service Tool/Quay/Quay Rhel/Rhelai3/Bootc Aws Cuda Rhel/Rhelai3/Bootc Azure Cuda Rhel/Rhelai3/Bootc Azure Rocm Rhel/Rhelai3/Bootc Gcp Cuda Rhel builds), which initialize _n_frames in __init__ instead of resetting it in _open(), and avoid re-setting _n_frames and is_animated in _seek() if already set, closing crafted Microsoft Image Composer (MIC) file with an embedded TIFF containing three or more IFDs. Until patched, this is a concern on Applications running Pillow 7.2.0 through 12.3.0 with olefile installed that query n_frames or extract frames from untrusted input, so Reject or sanitize untrusted MIC file uploads at ingestion and avoid calling n_frames or save_all on unverified images in environments where olefile is loaded. As a compensating measure, Impose CPU time limits using ulimit or prlimit to terminate worker processes caught in the non-advancing MIC image parsing loop before system CPU is exhausted and Configure service watchdogs with automatic restart under systemd to detect stalled processes via keep-alives and restore availability. These controls limit blast radius and contain exploitation, but they do not fix the underlying non-advancing frame traversal defect — apply the upstream patch to fully remediate.\n\nHighest-impact CME: CME-114 Process Resource Limits (ulimit/prlimit) (medium)","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"scores":[{"cvss_v3":{"attackComplexity":"LOW","attackVector":"NETWORK","availabilityImpact":"HIGH","baseScore":7.5,"baseSeverity":"HIGH","confidentialityImpact":"NONE","integrityImpact":"NONE","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"NONE","vectorString":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Important"}],"title":"pillow: Pillow: Denial of Service via infinite loop during crafted MIC image parsing"},{"cve":"CVE-2026-112391","cwe":{"id":"CWE-770","name":"Allocation of Resources Without Limits or Throttling"},"discovery_date":"2026-10-19T20:52:51.473281+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2555851"}],"notes":[{"category":"description","text":"A flaw was found in Pillow. The PDF parser fails to limit recursion depth when reading nested PDF data structures such as arrays and dictionaries. A remote attacker could exploit this vulnerability by supplying a specially crafted, deeply nested PDF file to an application processing PDFs in append mode. This causes excessive recursion that crashes or terminates the worker process, resulting in a Denial of Service (DoS).","title":"Vulnerability description"},{"category":"summary","text":"pillow: Pillow: Denial of Service via unbounded recursion in PDF parser","title":"Vulnerability summary"},{"category":"other","text":"Red Hat has evaluated this vulnerability as having Important severity because an unauthenticated remote attacker could cause worker process termination and denial of service by submitting maliciously structured input files. However, exploitation is constrained to specific deployment workflows where an application utilizes Pillow to append image data onto preexisting, untrusted PDF documents. Systems and services that only process standard raster image formats or do not manipulate external PDF files in append mode are unaffected.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-112391"},{"category":"external","summary":"RHBZ#2555851","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2555851"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-112391","url":"https://www.cve.org/CVERecord?id=CVE-2026-112391"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-112391","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-112391"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/commit/94cb11c44ff7345414da17ebfea1598a529ad9cd","url":"https://github.com/python-pillow/Pillow/commit/94cb11c44ff7345414da17ebfea1598a529ad9cd"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0","url":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-254f-3936-288h","url":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-254f-3936-288h"}],"release_date":"2026-10-19T20:40:36.285000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"},{"category":"workaround","details":"--- CME Mitigation Guidance (2026-10-20) ---\nDisclaimer: The following mitigation solutions are provided as a courtesy, pending the application of an official patch. This content was generated utilizing AI tools in conjunction with data from https://cmetaxonomy.org. Customers are advised to independently evaluate and test these temporary mitigations within their own infrastructure.\n\nThe highest-value action is to enforce recursion depth limits in PdfParser when reading nested PDF data structures during append operations: upgrade to Pillow 13.0.0 (or the fixed Python Pillow/Python3.11 Pillow/Rhoai/Odh Openvino Model Server Rhel/Rhelai3/Bootc Cuda Rhel/Rhelai3/Bootc Gaudi Rhel/Rhelai3/Bootc Rocm Rhel/Rhelai3/Disk Image Cuda Rhel/Python3X Pillow/Rhaiis/Vllm Spyre Rhel/Rhoai/Odh Kserve Storage Initializer Rhel/Rhoai/Odh Kserve Agent Rhel/Rhoai/Odh Kserve Controller Rhel/Rhoai/Odh Kserve Router Rhel/Rhoai/Odh Feature Server Rhel/Rhoai/Odh Trustyai Nemo Guardrails Server Rhel/Rhaiis/Vllm Tpu Rhel/Rhoai/Odh Training Cuda128 Torch29 Py312 Rhel/Rhoai/Odh Training Rocm64 Torch29 Py312 Rhel/Rhaiis/Vllm Cpu Rhel/Rhaiis/Vllm Neuron Rhel/Ansible Automation Platform 25/Lightspeed Chatbot Rhel/Openshift Lightspeed/Lightspeed Ocp Rag Rhel/Rhoai/Odh Th06 Cpu Torch291 Py312 Rhel/Rhoai/Odh Th06 Cuda130 Torch291 Py312 Rhel/Rhoai/Odh Vllm Gaudi Rhel/Python3.12 Pillow/Rhaii/Vllm Gaudi Rhel/Rhaii/Vllm Neuron Rhel/Rhaii/Vllm Tpu Rhel/Lightspeed Core/Lightspeed Stack Rhel/Rhoai/Odh Kserve Autogluon Server Rhel/Rhai Early Access/Docling Sdk Cuda Rhel/Rhai Early Access/Docling Serve Cuda Rhel/Rhoai/Odh Llm D Kv Cache Rhel/Rhoai/Odh Th06 Cpu Torch210 Py312 Rhel/Rhoai/Odh Th06 Cuda130 Torch210 Py312 Rhel/Rhoai/Odh Th06 Rocm64 Torch291 Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Llmcompressor Cuda Py312 Rhel/Ask Sre/Rhel Cla/Rlsapi Rhel/Rhaiis/Model Opt Cuda Rhel/Rhaiis/Vllm Cuda Rhel/Rhaiis/Vllm Rocm Rhel/Rhaii/Model Opt Cuda Rhel/Rhaii/Vllm Cpu Rhel/Rhaii/Vllm Cuda Rhel/Rhaii/Vllm Rocm Rhel/Rhaii/Vllm Spyre Rhel/Ansible Automation Platform/Automation Dashboard Rhel/Ansible Automation Platform 26/Hub Rhel/Ansible Automation Platform 26/Lightspeed Chatbot Rhel/Ansible Automation Platform 27/Hub Rhel/Ansible Automation Platform 27/Lightspeed Chatbot Rhel/Galaxy Ng/Automation Hub Container/Aap Rag Content/Ansible Chatbot Stack/Koku Report Emailer/Exploit Intelligence/Vulnerability Analysis Rhel/Insights Host Inventory/Google Lightspeed Agent/Lscore Deploy/Lightspeed Core/Rag Tool Cpu Rhel/Lightspeed Core/Rag Tool Cuda 12.9 Rhel/Lightspeed Services Ocp/Ocp Rules Rhel/Openshift Lightspeed Tech Preview/Lightspeed Rag Tool Rhel/Openshift Lightspeed/Lightspeed Service Api Rhel/Rhoai/Odh Caikit Tgis Serving Rhel/Rhoai/Odh Llama Stack Core Rhel/Rhoai/Odh Pipeline Runtime Pytorch Llmcompressor Cuda Py312 Rhel/Rhoai/Odh Automl Rhel/Rhoai/Odh Autorag Rhel/Rhoai/Odh Mlflow Rhel/Rhoai/Odh Pipeline Runtime Datascience Cpu Py312 Rhel/Rhoai/Odh Pipeline Runtime Pytorch Cuda Py312 Rhel/Rhoai/Odh Pipeline Runtime Pytorch Rocm Py312 Rhel/Rhoai/Odh Pipeline Runtime Tensorflow Cuda Py312 Rhel/Rhoai/Odh Pipeline Runtime Tensorflow Rocm Py312 Rhel/Rhoai/Odh Trustyai Garak Lls Provider Dsp Rhel/Rhoai/Odh Workbench Codeserver Datascience Cpu Py312 Rhel/Rhoai/Odh Workbench Jupyter Datascience Cpu Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Cuda Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Rocm Py312 Rhel/Rhoai/Odh Workbench Jupyter Tensorflow Cuda Py312 Rhel/Rhoai/Odh Workbench Jupyter Tensorflow Rocm Py312 Rhel/Rhoai/Odh Workbench Jupyter Trustyai Cpu Py312 Rhel/Rhai/Guidellm Rhel/Rhoai/Odh Ogx Core Rhel/Rhoai/Odh Th Torch Cpu Py312 Rhel/Rhoai/Odh Th Torch Cuda Py312 Rhel/Rhoai/Odh Th Torch Rocm Py312 Rhel/Quay/Quay Service Tool/Quay/Quay Rhel/Rhelai3/Bootc Aws Cuda Rhel/Rhelai3/Bootc Azure Cuda Rhel/Rhelai3/Bootc Azure Rocm Rhel/Rhelai3/Bootc Gcp Cuda Rhel builds), which set default max_nesting parameter to 900 instead of -1 in get_value() and read_indirect() to enforce recursion limits, closing all 2 injection vectors (deeply nested PDF arrays, deeply nested PDF dictionaries). Until patched, this is a concern on applications using Pillow to append images to existing, untrusted PDF files, so prevent untrusted users from submitting PDF files to applications executing image saves with PDF append enabled. As a compensating measure, configure service watchdogs and automatic restart policies such as systemd WatchdogSec and Restart to restart workers that terminate due to uncaught RecursionError exceptions. These controls limit blast radius and contain exploitation, but they do not fix the underlying unbounded recursion defect — apply the upstream patch to fully remediate.\n\nHighest-impact CME: CME-1204 Service Watchdog and Automatic Restart (systemd WatchdogSec/Restart) (medium)","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"scores":[{"cvss_v3":{"attackComplexity":"LOW","attackVector":"NETWORK","availabilityImpact":"HIGH","baseScore":7.5,"baseSeverity":"HIGH","confidentialityImpact":"NONE","integrityImpact":"NONE","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"NONE","vectorString":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Important"}],"title":"pillow: Pillow: Denial of Service via unbounded recursion in PDF parser"},{"cve":"CVE-2026-112392","cwe":{"id":"CWE-125","name":"Out-of-bounds Read"},"discovery_date":"2026-10-19T21:21:52.403747+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2555866"}],"notes":[{"category":"description","text":"A flaw was found in Pillow. An attacker can cause a Denial of Service (DoS) by submitting a zero-size image to an application that applies rank-based filters, such as median, minimum, or maximum filters. When attempting to expand the image borders, Pillow reads from unallocated image storage, resulting in an application crash.","title":"Vulnerability description"},{"category":"summary","text":"pillow: pillow: Denial of Service via zero-size images in rank filters","title":"Vulnerability summary"},{"category":"other","text":"Red Hat has rated this vulnerability as having an impact of Important because an unauthenticated remote attacker can crash backend application processes handling image transformations without requiring elevated privileges. In Red Hat environments deploying services with Pillow, processing untrusted image inputs or crop coordinates using rank-derived filtering routines can lead to service interruption. However, the scope of impact is confined strictly to process termination resulting from an invalid memory dereference, with no evidence of memory corruption or confidentiality exposure.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-112392"},{"category":"external","summary":"RHBZ#2555866","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2555866"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-112392","url":"https://www.cve.org/CVERecord?id=CVE-2026-112392"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-112392","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-112392"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/commit/443e3130a8c11594fd0e296afad497e11e678961","url":"https://github.com/python-pillow/Pillow/commit/443e3130a8c11594fd0e296afad497e11e678961"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0","url":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-27h8-gf7j-v5hc","url":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-27h8-gf7j-v5hc"}],"release_date":"2026-10-19T21:17:18.908000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"},{"category":"workaround","details":"--- CME Mitigation Guidance (2026-10-20) ---\nDisclaimer: The following mitigation solutions are provided as a courtesy, pending the application of an official patch. This content was generated utilizing AI tools in conjunction with data from https://cmetaxonomy.org. Customers are advised to independently evaluate and test these temporary mitigations within their own infrastructure.\n\nThe highest-value action is to validate image dimensions and reject zero-width or zero-height inputs before invoking rank, median, minimum, or maximum filters in Pillow: upgrade to Pillow 13.0.0 (or the fixed Python Pillow/Python3.11 Pillow/Rhoai/Odh Openvino Model Server Rhel/Rhelai3/Bootc Cuda Rhel/Rhelai3/Bootc Gaudi Rhel/Rhelai3/Bootc Rocm Rhel/Rhelai3/Disk Image Cuda Rhel/Python3X Pillow/Rhaiis/Vllm Spyre Rhel/Rhoai/Odh Kserve Storage Initializer Rhel/Rhoai/Odh Kserve Agent Rhel/Rhoai/Odh Kserve Controller Rhel/Rhoai/Odh Kserve Router Rhel/Rhoai/Odh Feature Server Rhel/Rhoai/Odh Trustyai Nemo Guardrails Server Rhel/Rhaiis/Vllm Tpu Rhel/Rhoai/Odh Training Cuda128 Torch29 Py312 Rhel/Rhoai/Odh Training Rocm64 Torch29 Py312 Rhel/Rhaiis/Vllm Cpu Rhel/Rhaiis/Vllm Neuron Rhel/Ansible Automation Platform 25/Lightspeed Chatbot Rhel/Openshift Lightspeed/Lightspeed Ocp Rag Rhel/Rhoai/Odh Th06 Cpu Torch291 Py312 Rhel/Rhoai/Odh Th06 Cuda130 Torch291 Py312 Rhel/Rhoai/Odh Vllm Gaudi Rhel/Python3.12 Pillow/Rhaii/Vllm Gaudi Rhel/Rhaii/Vllm Neuron Rhel/Rhaii/Vllm Tpu Rhel/Lightspeed Core/Lightspeed Stack Rhel/Rhoai/Odh Kserve Autogluon Server Rhel/Rhai Early Access/Docling Sdk Cuda Rhel/Rhai Early Access/Docling Serve Cuda Rhel/Rhoai/Odh Llm D Kv Cache Rhel/Rhoai/Odh Th06 Cpu Torch210 Py312 Rhel/Rhoai/Odh Th06 Cuda130 Torch210 Py312 Rhel/Rhoai/Odh Th06 Rocm64 Torch291 Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Llmcompressor Cuda Py312 Rhel/Ask Sre/Rhel Cla/Rlsapi Rhel/Rhaiis/Model Opt Cuda Rhel/Rhaiis/Vllm Cuda Rhel/Rhaiis/Vllm Rocm Rhel/Rhaii/Model Opt Cuda Rhel/Rhaii/Vllm Cpu Rhel/Rhaii/Vllm Cuda Rhel/Rhaii/Vllm Rocm Rhel/Rhaii/Vllm Spyre Rhel/Ansible Automation Platform/Automation Dashboard Rhel/Ansible Automation Platform 26/Hub Rhel/Ansible Automation Platform 26/Lightspeed Chatbot Rhel/Ansible Automation Platform 27/Hub Rhel/Ansible Automation Platform 27/Lightspeed Chatbot Rhel/Galaxy Ng/Automation Hub Container/Aap Rag Content/Ansible Chatbot Stack/Koku Report Emailer/Exploit Intelligence/Vulnerability Analysis Rhel/Insights Host Inventory/Google Lightspeed Agent/Lscore Deploy/Lightspeed Core/Rag Tool Cpu Rhel/Lightspeed Core/Rag Tool Cuda 12.9 Rhel/Lightspeed Services Ocp/Ocp Rules Rhel/Openshift Lightspeed Tech Preview/Lightspeed Rag Tool Rhel/Openshift Lightspeed/Lightspeed Service Api Rhel/Rhoai/Odh Caikit Tgis Serving Rhel/Rhoai/Odh Llama Stack Core Rhel/Rhoai/Odh Pipeline Runtime Pytorch Llmcompressor Cuda Py312 Rhel/Rhoai/Odh Automl Rhel/Rhoai/Odh Autorag Rhel/Rhoai/Odh Mlflow Rhel/Rhoai/Odh Pipeline Runtime Datascience Cpu Py312 Rhel/Rhoai/Odh Pipeline Runtime Pytorch Cuda Py312 Rhel/Rhoai/Odh Pipeline Runtime Pytorch Rocm Py312 Rhel/Rhoai/Odh Pipeline Runtime Tensorflow Cuda Py312 Rhel/Rhoai/Odh Pipeline Runtime Tensorflow Rocm Py312 Rhel/Rhoai/Odh Trustyai Garak Lls Provider Dsp Rhel/Rhoai/Odh Workbench Codeserver Datascience Cpu Py312 Rhel/Rhoai/Odh Workbench Jupyter Datascience Cpu Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Cuda Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Rocm Py312 Rhel/Rhoai/Odh Workbench Jupyter Tensorflow Cuda Py312 Rhel/Rhoai/Odh Workbench Jupyter Tensorflow Rocm Py312 Rhel/Rhoai/Odh Workbench Jupyter Trustyai Cpu Py312 Rhel/Rhai/Guidellm Rhel/Rhoai/Odh Ogx Core Rhel/Rhoai/Odh Th Torch Cpu Py312 Rhel/Rhoai/Odh Th Torch Cuda Py312 Rhel/Rhoai/Odh Th Torch Rocm Py312 Rhel/Quay/Quay Service Tool/Quay/Quay Rhel/Rhelai3/Bootc Aws Cuda Rhel/Rhelai3/Bootc Azure Cuda Rhel/Rhelai3/Bootc Azure Rocm Rhel/Rhelai3/Bootc Gcp Cuda Rhel builds), which check if imIn->xsize or imIn->ysize is zero in ImagingExpand(), allocate a new image, copy palette data, and return early without performing line expansion, closing all 3 injection vectors (multi-page TIFFs with zero-dimension secondary IFDs, image crops with identical edge boundaries resulting in zero dimensions, zero-width or zero-height input images). Until patched, this is a concern on Applications accepting untrusted images or crop parameters that process zero-dimension images through rank-based filters, so Validate and reject zero-dimension images or identical crop coordinates prior to passing image objects to rank-based filtering routines. These controls limit blast radius and contain exploitation, but they do not fix the underlying NULL pointer dereference defect — apply the upstream patch to fully remediate.","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"scores":[{"cvss_v3":{"attackComplexity":"LOW","attackVector":"NETWORK","availabilityImpact":"HIGH","baseScore":7.5,"baseSeverity":"HIGH","confidentialityImpact":"NONE","integrityImpact":"NONE","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"NONE","vectorString":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Important"}],"title":"pillow: pillow: Denial of Service via zero-size images in rank filters"},{"cve":"CVE-2026-112393","cwe":{"id":"CWE-787","name":"Out-of-bounds Write"},"discovery_date":"2026-10-19T20:51:29.035172+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2555846"}],"notes":[{"category":"description","text":"A flaw was found in Pillow. When saving compressed Tagged Image File Format (TIFF) images into memory buffers rather than direct files, the library incorrectly calculates the required memory allocation. A remote attacker who supplies a crafted image for processing can trigger a heap-based buffer overflow, potentially leading to arbitrary code execution or a Denial of Service (DoS) through process crashes. Additionally, uninitialized memory gaps can result in information disclosure by leaking remnants of previously stored data.","title":"Vulnerability description"},{"category":"summary","text":"pillow: Pillow: Heap buffer overflow in _tiffWriteProc when saving TIFF to BytesIO","title":"Vulnerability summary"},{"category":"other","text":"The attack complexity is high and also for services that do hit this code path (image upload → TIFF re-encode to BytesIO → return/upload), the minimal PoC reliably crashes the worker process. The heap corruption could plausibly be developed into RCE by a skilled attacker who has already compromised the same process enough to leak addresses, but that's a chained-exploit scenario, not a standalone wormable RCE.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-112393"},{"category":"external","summary":"RHBZ#2555846","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2555846"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-112393","url":"https://www.cve.org/CVERecord?id=CVE-2026-112393"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-112393","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-112393"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/commit/11582e5ffbdceb973d5ee593f3105af5a9852b40","url":"https://github.com/python-pillow/Pillow/commit/11582e5ffbdceb973d5ee593f3105af5a9852b40"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0","url":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-332p-7hrc-92wf","url":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-332p-7hrc-92wf"}],"release_date":"2026-10-19T20:46:40.784000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"}],"scores":[{"cvss_v3":{"attackComplexity":"HIGH","attackVector":"NETWORK","availabilityImpact":"HIGH","baseScore":8.1,"baseSeverity":"HIGH","confidentialityImpact":"HIGH","integrityImpact":"HIGH","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"NONE","vectorString":"CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:H/I:H/A:H","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Important"}],"title":"pillow: Pillow: Heap buffer overflow in _tiffWriteProc when saving TIFF to BytesIO"},{"cve":"CVE-2026-112394","cwe":{"id":"CWE-125","name":"Out-of-bounds Read"},"discovery_date":"2026-10-19T20:51:16.036574+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2555845"}],"notes":[{"category":"description","text":"A flaw was found in pillow. A remote attacker could trigger an out-of-bounds read by supplying specially crafted image dimensions and input buffers to the raw image decoders. Because the decoder calculates an image row size smaller than the unpacker's actual memory footprint, the library reads past the end of the buffer. This flaw can lead to information disclosure by exposing adjacent heap memory in output pixels or cause a Denial of Service (DoS) by crashing the application.","title":"Vulnerability description"},{"category":"summary","text":"pillow: pillow: Information disclosure via out-of-bounds read in raw decoders","title":"Vulnerability summary"},{"category":"other","text":"This vulnerability is rated as Moderate severity because successful exploitation requires an application to directly invoke raw decoding interfaces on untrusted input using specific planar bit modes. Standard file format decoders bundled in Pillow do not trigger this condition during normal image loading operations. Additionally, the potential impact is constrained to process crashes or disclosing at most three neighboring heap bytes, without allowing arbitrary code execution.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-112394"},{"category":"external","summary":"RHBZ#2555845","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2555845"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-112394","url":"https://www.cve.org/CVERecord?id=CVE-2026-112394"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-112394","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-112394"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/commit/392ee6f0d19a4adb766a63e9a99f3c28193b8d31","url":"https://github.com/python-pillow/Pillow/commit/392ee6f0d19a4adb766a63e9a99f3c28193b8d31"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0","url":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-4558-3h8r-v8xp","url":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-4558-3h8r-v8xp"}],"release_date":"2026-10-19T20:48:28.141000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"}],"scores":[{"cvss_v3":{"attackComplexity":"LOW","attackVector":"NETWORK","availabilityImpact":"LOW","baseScore":6.5,"baseSeverity":"MEDIUM","confidentialityImpact":"LOW","integrityImpact":"NONE","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"NONE","vectorString":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:N/A:L","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Moderate"}],"title":"pillow: pillow: Information disclosure via out-of-bounds read in raw decoders"},{"cve":"CVE-2026-112395","cwe":{"id":"CWE-125","name":"Out-of-bounds Read"},"discovery_date":"2026-10-19T21:11:16.697842+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2555857"}],"notes":[{"category":"description","text":"A flaw was found in Pillow. Its tiled-TIFF image decoder does not properly restrict buffer boundaries when copying pixel data, leading to an out-of-bounds read. A remote attacker could exploit this vulnerability by providing a specially crafted tiled TIFF image to an application using the library. Successful exploitation could allow an attacker to obtain sensitive information from adjacent heap memory or cause a Denial of Service (DoS) by crashing the application.","title":"Vulnerability description"},{"category":"summary","text":"pillow: pillow: Information disclosure via crafted tiled TIFF image","title":"Vulnerability summary"},{"category":"other","text":"This vulnerability is evaluated as having a Moderate impact because successful exploitation relies on an application ingesting and parsing an untrusted, specifically formatted tiled TIFF file. While reading outside the allocated tile buffer allows exposure of adjacent process memory within reconstructed image data or can induce an application crash, it does not permit arbitrary remote code execution or modification of system files. Red Hat services and container workloads that do not parse external user-supplied TIFF images remain unaffected in practice.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-112395"},{"category":"external","summary":"RHBZ#2555857","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2555857"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-112395","url":"https://www.cve.org/CVERecord?id=CVE-2026-112395"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-112395","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-112395"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/commit/76869efdbbf2ad7dccd7536520fb61b7a00ab61e","url":"https://github.com/python-pillow/Pillow/commit/76869efdbbf2ad7dccd7536520fb61b7a00ab61e"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0","url":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-462x-pgg6-55fp","url":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-462x-pgg6-55fp"}],"release_date":"2026-10-19T20:59:15.098000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"}],"scores":[{"cvss_v3":{"attackComplexity":"LOW","attackVector":"NETWORK","availabilityImpact":"LOW","baseScore":6.5,"baseSeverity":"MEDIUM","confidentialityImpact":"LOW","integrityImpact":"NONE","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"NONE","vectorString":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:N/A:L","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Moderate"}],"title":"pillow: pillow: Information disclosure via crafted tiled TIFF image"},{"cve":"CVE-2026-112396","cwe":{"id":"CWE-825","name":"Expired Pointer Dereference"},"discovery_date":"2026-10-19T21:21:47.568743+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2555865"}],"notes":[{"category":"description","text":"A flaw was found in Pillow. This vulnerability allows a remote attacker to potentially execute arbitrary code or cause a Denial of Service (DoS). The issue occurs when an application encodes Tagged Image File Format (TIFF) data with untrusted tags into an in-memory buffer. When an invalid tag is rejected during encoding, improper error handling prematurely frees memory that is subsequently accessed during cleanup, triggering a heap use-after-free condition.","title":"Vulnerability description"},{"category":"summary","text":"pillow: pillow: arbitrary code execution via heap use-after-free during TIFF encoding","title":"Vulnerability summary"},{"category":"other","text":"This vulnerability is rated as Moderate impact for Red Hat products because exploitation requires specific and uncommon application workflows where untrusted metadata is accepted and serialized into TIFF format within an in-memory buffer. Exploitation relies on strict preconditions, specifically libtiff rejecting malformed tags during in-memory encoding, making remote code execution difficult and predominantly resulting in process termination. Typical Red Hat product deployments do not expose arbitrary, in-memory TIFF re-encoding endpoints accepting unvalidated user tags directly to unauthenticated network clients.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-112396"},{"category":"external","summary":"RHBZ#2555865","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2555865"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-112396","url":"https://www.cve.org/CVERecord?id=CVE-2026-112396"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-112396","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-112396"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/commit/dc95da20207827a062cfbf551ee9f6031465a8bb","url":"https://github.com/python-pillow/Pillow/commit/dc95da20207827a062cfbf551ee9f6031465a8bb"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0","url":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-4rr5-89j8-mwwm","url":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-4rr5-89j8-mwwm"}],"release_date":"2026-10-19T21:02:15.197000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"}],"scores":[{"cvss_v3":{"attackComplexity":"HIGH","attackVector":"NETWORK","availabilityImpact":"HIGH","baseScore":5.9,"baseSeverity":"MEDIUM","confidentialityImpact":"NONE","integrityImpact":"NONE","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"NONE","vectorString":"CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:N/A:H","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Moderate"}],"title":"pillow: pillow: arbitrary code execution via heap use-after-free during TIFF encoding"},{"cve":"CVE-2026-112397","cwe":{"id":"CWE-124","name":"Buffer Underwrite"},"discovery_date":"2026-10-19T21:22:05.164607+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2555868"}],"notes":[{"category":"description","text":"A flaw was found in Pillow. A remote attacker can cause a Denial of Service (DoS) by supplying a specially crafted image with a width of zero to an application performing morphology operations. Because Pillow fails to validate zero-width image geometry during these operations, it attempts an out-of-bounds memory access before the buffer boundary, resulting in an application crash.","title":"Vulnerability description"},{"category":"summary","text":"pillow: pillow: Denial of Service via out-of-bounds memory access on zero-width images","title":"Vulnerability summary"},{"category":"other","text":"This vulnerability is rated as having Important impact because remote, unauthenticated attackers can trigger an immediate denial of service by crashing target processes without requiring user intervention. In typical Red Hat product deployments, exposure is limited because triggering the flaw requires an application to actively process untrusted image input through morphological transformations. Workloads that parse standard image formats without performing morphology operations are not susceptible to this failure mode.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-112397"},{"category":"external","summary":"RHBZ#2555868","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2555868"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-112397","url":"https://www.cve.org/CVERecord?id=CVE-2026-112397"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-112397","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-112397"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/commit/18f7c0dfee93063c51db6ed0ca8c4da2999dcb64","url":"https://github.com/python-pillow/Pillow/commit/18f7c0dfee93063c51db6ed0ca8c4da2999dcb64"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0","url":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-58r4-6rv3-gfh8","url":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-58r4-6rv3-gfh8"}],"release_date":"2026-10-19T21:17:35.684000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"},{"category":"workaround","details":"--- CME Mitigation Guidance (2026-10-20) ---\nDisclaimer: The following mitigation solutions are provided as a courtesy, pending the application of an official patch. This content was generated utilizing AI tools in conjunction with data from https://cmetaxonomy.org. Customers are advised to independently evaluate and test these temporary mitigations within their own infrastructure.\n\nThe highest-value action is to validate image dimensions to reject zero-width or zero-height inputs before invoking ImageMorph.MorphOp.apply() in Python Pillow: upgrade to Pillow 13.0.0 (or the fixed Python Pillow/Python3.11 Pillow/Rhoai/Odh Openvino Model Server Rhel/Rhelai3/Bootc Cuda Rhel/Rhelai3/Bootc Gaudi Rhel/Rhelai3/Bootc Rocm Rhel/Rhelai3/Disk Image Cuda Rhel/Python3X Pillow/Rhaiis/Vllm Spyre Rhel/Rhoai/Odh Kserve Storage Initializer Rhel/Rhoai/Odh Kserve Agent Rhel/Rhoai/Odh Kserve Controller Rhel/Rhoai/Odh Kserve Router Rhel/Rhoai/Odh Feature Server Rhel/Rhoai/Odh Trustyai Nemo Guardrails Server Rhel/Rhaiis/Vllm Tpu Rhel/Rhoai/Odh Training Cuda128 Torch29 Py312 Rhel/Rhoai/Odh Training Rocm64 Torch29 Py312 Rhel/Rhaiis/Vllm Cpu Rhel/Rhaiis/Vllm Neuron Rhel/Ansible Automation Platform 25/Lightspeed Chatbot Rhel/Openshift Lightspeed/Lightspeed Ocp Rag Rhel/Rhoai/Odh Th06 Cpu Torch291 Py312 Rhel/Rhoai/Odh Th06 Cuda130 Torch291 Py312 Rhel/Rhoai/Odh Vllm Gaudi Rhel/Python3.12 Pillow/Rhaii/Vllm Gaudi Rhel/Rhaii/Vllm Neuron Rhel/Rhaii/Vllm Tpu Rhel/Lightspeed Core/Lightspeed Stack Rhel/Rhoai/Odh Kserve Autogluon Server Rhel/Rhai Early Access/Docling Sdk Cuda Rhel/Rhai Early Access/Docling Serve Cuda Rhel/Rhoai/Odh Llm D Kv Cache Rhel/Rhoai/Odh Th06 Cpu Torch210 Py312 Rhel/Rhoai/Odh Th06 Cuda130 Torch210 Py312 Rhel/Rhoai/Odh Th06 Rocm64 Torch291 Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Llmcompressor Cuda Py312 Rhel/Ask Sre/Rhel Cla/Rlsapi Rhel/Rhaiis/Model Opt Cuda Rhel/Rhaiis/Vllm Cuda Rhel/Rhaiis/Vllm Rocm Rhel/Rhaii/Model Opt Cuda Rhel/Rhaii/Vllm Cpu Rhel/Rhaii/Vllm Cuda Rhel/Rhaii/Vllm Rocm Rhel/Rhaii/Vllm Spyre Rhel/Ansible Automation Platform/Automation Dashboard Rhel/Ansible Automation Platform 26/Hub Rhel/Ansible Automation Platform 26/Lightspeed Chatbot Rhel/Ansible Automation Platform 27/Hub Rhel/Ansible Automation Platform 27/Lightspeed Chatbot Rhel/Galaxy Ng/Automation Hub Container/Aap Rag Content/Ansible Chatbot Stack/Koku Report Emailer/Exploit Intelligence/Vulnerability Analysis Rhel/Insights Host Inventory/Google Lightspeed Agent/Lscore Deploy/Lightspeed Core/Rag Tool Cpu Rhel/Lightspeed Core/Rag Tool Cuda 12.9 Rhel/Lightspeed Services Ocp/Ocp Rules Rhel/Openshift Lightspeed Tech Preview/Lightspeed Rag Tool Rhel/Openshift Lightspeed/Lightspeed Service Api Rhel/Rhoai/Odh Caikit Tgis Serving Rhel/Rhoai/Odh Llama Stack Core Rhel/Rhoai/Odh Pipeline Runtime Pytorch Llmcompressor Cuda Py312 Rhel/Rhoai/Odh Automl Rhel/Rhoai/Odh Autorag Rhel/Rhoai/Odh Mlflow Rhel/Rhoai/Odh Pipeline Runtime Datascience Cpu Py312 Rhel/Rhoai/Odh Pipeline Runtime Pytorch Cuda Py312 Rhel/Rhoai/Odh Pipeline Runtime Pytorch Rocm Py312 Rhel/Rhoai/Odh Pipeline Runtime Tensorflow Cuda Py312 Rhel/Rhoai/Odh Pipeline Runtime Tensorflow Rocm Py312 Rhel/Rhoai/Odh Trustyai Garak Lls Provider Dsp Rhel/Rhoai/Odh Workbench Codeserver Datascience Cpu Py312 Rhel/Rhoai/Odh Workbench Jupyter Datascience Cpu Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Cuda Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Rocm Py312 Rhel/Rhoai/Odh Workbench Jupyter Tensorflow Cuda Py312 Rhel/Rhoai/Odh Workbench Jupyter Tensorflow Rocm Py312 Rhel/Rhoai/Odh Workbench Jupyter Trustyai Cpu Py312 Rhel/Rhai/Guidellm Rhel/Rhoai/Odh Ogx Core Rhel/Rhoai/Odh Th Torch Cpu Py312 Rhel/Rhoai/Odh Th Torch Cuda Py312 Rhel/Rhoai/Odh Th Torch Rocm Py312 Rhel/Quay/Quay Service Tool/Quay/Quay Rhel/Rhelai3/Bootc Aws Cuda Rhel/Rhelai3/Bootc Azure Cuda Rhel/Rhelai3/Bootc Azure Rocm Rhel/Rhelai3/Bootc Gcp Cuda Rhel builds), which check for zero width or height and return early before executing morphology row manipulations in src/_imagingmorph.c, closing all 2 injection vectors (crafted zero-width PPM image, crafted zero-width PGM image). Until patched, this is a concern on Applications processing untrusted images using Pillow's ImageMorph.MorphOp.apply(), so Reject incoming image files with zero width or zero height, particularly PPM and PGM formats, before submitting them to morphology routines. These controls limit blast radius and contain exploitation, but they do not fix the underlying out-of-bounds memory access buffer underflow defect — apply the upstream patch to fully remediate.","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"scores":[{"cvss_v3":{"attackComplexity":"LOW","attackVector":"NETWORK","availabilityImpact":"HIGH","baseScore":7.5,"baseSeverity":"HIGH","confidentialityImpact":"NONE","integrityImpact":"NONE","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"NONE","vectorString":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Important"}],"title":"pillow: pillow: Denial of Service via out-of-bounds memory access on zero-width images"},{"cve":"CVE-2026-112398","cwe":{"id":"CWE-787","name":"Out-of-bounds Write"},"discovery_date":"2026-10-19T21:21:59.767228+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2555867"}],"notes":[{"category":"description","text":"A flaw was found in Pillow. A remote attacker can achieve arbitrary code execution by supplying a specially crafted FlashPix (FPX) image file to an application that processes untrusted inputs. Due to improper handling of image tile data during JPEG decoding, the library writes pixel data beyond the allocated memory buffer. This out-of-bounds write corrupts heap memory and can allow an attacker to execute arbitrary commands or cause an application crash.","title":"Vulnerability description"},{"category":"summary","text":"pillow: Pillow: Heap buffer overflow in FlashPix (FPX) JPEG decoding leads to code execution","title":"Vulnerability summary"},{"category":"other","text":"- Requires optional olefile package installed (not bundled by default in RHEL/most containers)\n- High attack complexity for reliable RCE (heap layout control, vtable hijacking)\n- Affects a file format (FlashPix) rarely used in production image processing","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-112398"},{"category":"external","summary":"RHBZ#2555867","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2555867"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-112398","url":"https://www.cve.org/CVERecord?id=CVE-2026-112398"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-112398","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-112398"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/commit/242a059aedc1935097d9aa1dc3f099750467ba75","url":"https://github.com/python-pillow/Pillow/commit/242a059aedc1935097d9aa1dc3f099750467ba75"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0","url":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-5x8w-8jj5-q52h","url":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-5x8w-8jj5-q52h"}],"release_date":"2026-10-19T21:07:25.229000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"}],"scores":[{"cvss_v3":{"attackComplexity":"HIGH","attackVector":"NETWORK","availabilityImpact":"HIGH","baseScore":8.1,"baseSeverity":"HIGH","confidentialityImpact":"HIGH","integrityImpact":"HIGH","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"NONE","vectorString":"CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:H/I:H/A:H","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Important"}],"title":"pillow: Pillow: Heap buffer overflow in FlashPix (FPX) JPEG decoding leads to code execution"},{"cve":"CVE-2026-112399","cwe":{"id":"CWE-770","name":"Allocation of Resources Without Limits or Throttling"},"discovery_date":"2026-10-19T21:21:30.155551+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2555863"}],"notes":[{"category":"description","text":"A flaw was found in Pillow. A remote attacker can trigger a Denial of Service (DoS) by providing a specially crafted Animated Portable Network Graphics (APNG) image. Pillow allocates memory for the full canvas dimensions before validating safety size limits during initial file opening. Processing such an image can exhaust available system memory and cause the application to crash.","title":"Vulnerability description"},{"category":"summary","text":"pillow: pillow: Denial of service via crafted APNG image","title":"Vulnerability summary"},{"category":"other","text":"This vulnerability is rated as an Important severity for Red Hat products because an unauthenticated remote attacker can cause a denial of service through memory exhaustion without requiring user interaction or elevated privileges. Red Hat solutions such as OpenShift AI and Ansible Automation Platform integrate Pillow to process arbitrary user or automated media uploads. Because full-canvas memory allocation occurs during initial file header ingestion prior to decompression safety limits being enforced, applications cannot rely on standard dimension safeguards to prevent worker termination.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-112399"},{"category":"external","summary":"RHBZ#2555863","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2555863"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-112399","url":"https://www.cve.org/CVERecord?id=CVE-2026-112399"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-112399","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-112399"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/commit/f1acfc54c53502948d0d14121872061df931db48","url":"https://github.com/python-pillow/Pillow/commit/f1acfc54c53502948d0d14121872061df931db48"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0","url":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-5xm2-6676-qcp9","url":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-5xm2-6676-qcp9"}],"release_date":"2026-10-19T21:10:11.266000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"},{"category":"workaround","details":"--- CME Mitigation Guidance (2026-10-20) ---\nDisclaimer: The following mitigation solutions are provided as a courtesy, pending the application of an official patch. This content was generated utilizing AI tools in conjunction with data from https://cmetaxonomy.org. Customers are advised to independently evaluate and test these temporary mitigations within their own infrastructure.\n\nThe highest-value action is to enforce decompression bomb validation checks prior to full-canvas memory allocation during initial image parsing in Pillow: upgrade to Pillow 13.0.0 (or the fixed Python Pillow/Python3.11 Pillow/Rhoai/Odh Openvino Model Server Rhel/Rhelai3/Bootc Cuda Rhel/Rhelai3/Bootc Gaudi Rhel/Rhelai3/Bootc Rocm Rhel/Rhelai3/Disk Image Cuda Rhel/Python3X Pillow/Rhaiis/Vllm Spyre Rhel/Rhoai/Odh Kserve Storage Initializer Rhel/Rhoai/Odh Kserve Agent Rhel/Rhoai/Odh Kserve Controller Rhel/Rhoai/Odh Kserve Router Rhel/Rhoai/Odh Feature Server Rhel/Rhoai/Odh Trustyai Nemo Guardrails Server Rhel/Rhaiis/Vllm Tpu Rhel/Rhoai/Odh Training Cuda128 Torch29 Py312 Rhel/Rhoai/Odh Training Rocm64 Torch29 Py312 Rhel/Rhaiis/Vllm Cpu Rhel/Rhaiis/Vllm Neuron Rhel/Ansible Automation Platform 25/Lightspeed Chatbot Rhel/Openshift Lightspeed/Lightspeed Ocp Rag Rhel/Rhoai/Odh Th06 Cpu Torch291 Py312 Rhel/Rhoai/Odh Th06 Cuda130 Torch291 Py312 Rhel/Rhoai/Odh Vllm Gaudi Rhel/Python3.12 Pillow/Rhaii/Vllm Gaudi Rhel/Rhaii/Vllm Neuron Rhel/Rhaii/Vllm Tpu Rhel/Lightspeed Core/Lightspeed Stack Rhel/Rhoai/Odh Kserve Autogluon Server Rhel/Rhai Early Access/Docling Sdk Cuda Rhel/Rhai Early Access/Docling Serve Cuda Rhel/Rhoai/Odh Llm D Kv Cache Rhel/Rhoai/Odh Th06 Cpu Torch210 Py312 Rhel/Rhoai/Odh Th06 Cuda130 Torch210 Py312 Rhel/Rhoai/Odh Th06 Rocm64 Torch291 Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Llmcompressor Cuda Py312 Rhel/Ask Sre/Rhel Cla/Rlsapi Rhel/Rhaiis/Model Opt Cuda Rhel/Rhaiis/Vllm Cuda Rhel/Rhaiis/Vllm Rocm Rhel/Rhaii/Model Opt Cuda Rhel/Rhaii/Vllm Cpu Rhel/Rhaii/Vllm Cuda Rhel/Rhaii/Vllm Rocm Rhel/Rhaii/Vllm Spyre Rhel/Ansible Automation Platform/Automation Dashboard Rhel/Ansible Automation Platform 26/Hub Rhel/Ansible Automation Platform 26/Lightspeed Chatbot Rhel/Ansible Automation Platform 27/Hub Rhel/Ansible Automation Platform 27/Lightspeed Chatbot Rhel/Galaxy Ng/Automation Hub Container/Aap Rag Content/Ansible Chatbot Stack/Koku Report Emailer/Exploit Intelligence/Vulnerability Analysis Rhel/Insights Host Inventory/Google Lightspeed Agent/Lscore Deploy/Lightspeed Core/Rag Tool Cpu Rhel/Lightspeed Core/Rag Tool Cuda 12.9 Rhel/Lightspeed Services Ocp/Ocp Rules Rhel/Openshift Lightspeed Tech Preview/Lightspeed Rag Tool Rhel/Openshift Lightspeed/Lightspeed Service Api Rhel/Rhoai/Odh Caikit Tgis Serving Rhel/Rhoai/Odh Llama Stack Core Rhel/Rhoai/Odh Pipeline Runtime Pytorch Llmcompressor Cuda Py312 Rhel/Rhoai/Odh Automl Rhel/Rhoai/Odh Autorag Rhel/Rhoai/Odh Mlflow Rhel/Rhoai/Odh Pipeline Runtime Datascience Cpu Py312 Rhel/Rhoai/Odh Pipeline Runtime Pytorch Cuda Py312 Rhel/Rhoai/Odh Pipeline Runtime Pytorch Rocm Py312 Rhel/Rhoai/Odh Pipeline Runtime Tensorflow Cuda Py312 Rhel/Rhoai/Odh Pipeline Runtime Tensorflow Rocm Py312 Rhel/Rhoai/Odh Trustyai Garak Lls Provider Dsp Rhel/Rhoai/Odh Workbench Codeserver Datascience Cpu Py312 Rhel/Rhoai/Odh Workbench Jupyter Datascience Cpu Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Cuda Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Rocm Py312 Rhel/Rhoai/Odh Workbench Jupyter Tensorflow Cuda Py312 Rhel/Rhoai/Odh Workbench Jupyter Tensorflow Rocm Py312 Rhel/Rhoai/Odh Workbench Jupyter Trustyai Cpu Py312 Rhel/Rhai/Guidellm Rhel/Rhoai/Odh Ogx Core Rhel/Rhoai/Odh Th Torch Cpu Py312 Rhel/Rhoai/Odh Th Torch Cuda Py312 Rhel/Rhoai/Odh Th Torch Rocm Py312 Rhel/Quay/Quay Service Tool/Quay/Quay Rhel/Rhelai3/Bootc Aws Cuda Rhel/Rhelai3/Bootc Azure Cuda Rhel/Rhelai3/Bootc Azure Rocm Rhel/Rhelai3/Bootc Gcp Cuda Rhel builds), which call Image._decompression_bomb_check(self.size) immediately after opening each frame to raise DecompressionBombError before allocating frame buffers, closing all 2 injection vectors (crafted APNG image with OP_BACKGROUND disposal, crafted APNG image with first-frame OP_PREVIOUS disposal). Until patched, this is a concern on Applications on any platform opening untrusted image files with Pillow Image.open, so Restrict ingestion of untrusted APNG files and validate image header metadata before passing input to Image.open. As a compensating measure, Monitor memory pressure events and PSI metrics to detect abnormal allocation spikes from oversized canvas sizing, Enforce process address space limits using ulimit or prlimit RLIMIT_AS to restrict maximum memory consumption during parsing, Configure systemd service watchdogs and automatic restarts to restore availability if process termination occurs, and Apply cgroups v2 memory constraints to isolate container or service memory exhaustion and protect host availability. These controls limit blast radius and contain exploitation, but they do not fix the underlying uncontrolled memory allocation defect — apply the upstream patch to fully remediate.\n\nHighest-impact CME: CME-1010 cgroup Memory Pressure Monitoring (memory.events / PSI) (medium)","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"scores":[{"cvss_v3":{"attackComplexity":"LOW","attackVector":"NETWORK","availabilityImpact":"HIGH","baseScore":7.5,"baseSeverity":"HIGH","confidentialityImpact":"NONE","integrityImpact":"NONE","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"NONE","vectorString":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Important"}],"title":"pillow: pillow: Denial of service via crafted APNG image"},{"cve":"CVE-2026-112400","cwe":{"id":"CWE-770","name":"Allocation of Resources Without Limits or Throttling"},"discovery_date":"2026-10-19T21:21:41.148327+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2555864"}],"notes":[{"category":"description","text":"A flaw was found in Pillow. When parsing a crafted GIF image, the library fails to discard rejected frame dimensions after encountering an excessively large image exception. If an application catches this exception and continues processing the image, subsequent operations attempt to allocate memory based on the oversized dimensions. A remote attacker can exploit this issue to exhaust system memory, resulting in a Denial of Service (DoS).","title":"Vulnerability description"},{"category":"summary","text":"pillow: pillow: Denial of Service via memory exhaustion from crafted GIF image","title":"Vulnerability summary"},{"category":"other","text":"Red Hat rates this vulnerability as Important because an unauthenticated remote attacker can supply a malicious graphic file that forces extensive memory allocations, leading to daemon termination or denial of service. The flaw requires consuming applications to catch decompression bomb exceptions and reuse the same image object across subsequent frame operations, allowing oversized geometry values to trigger high resource exhaustion. Exposure is largely concentrated in containerized services and image-processing pipelines that handle untrusted animation uploads directly.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-112400"},{"category":"external","summary":"RHBZ#2555864","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2555864"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-112400","url":"https://www.cve.org/CVERecord?id=CVE-2026-112400"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-112400","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-112400"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/commit/6faed707cfb760a4cbd726ddc8406f576cf35bf7","url":"https://github.com/python-pillow/Pillow/commit/6faed707cfb760a4cbd726ddc8406f576cf35bf7"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0","url":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-65qm-jvmh-6pqg","url":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-65qm-jvmh-6pqg"}],"release_date":"2026-10-19T21:12:37.401000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"},{"category":"workaround","details":"--- CME Mitigation Guidance (2026-10-20) ---\nDisclaimer: The following mitigation solutions are provided as a courtesy, pending the application of an official patch. This content was generated utilizing AI tools in conjunction with data from https://cmetaxonomy.org. Customers are advised to independently evaluate and test these temporary mitigations within their own infrastructure.\n\nThe highest-value action is to discard image objects immediately upon catching DecompressionBombError rather than continuing seek or load operations: upgrade to Pillow 13.0.0 (or the fixed Python Pillow/Python3.11 Pillow/Rhoai/Odh Openvino Model Server Rhel/Rhelai3/Bootc Cuda Rhel/Rhelai3/Bootc Gaudi Rhel/Rhelai3/Bootc Rocm Rhel/Rhelai3/Disk Image Cuda Rhel/Python3X Pillow/Rhaiis/Vllm Spyre Rhel/Rhoai/Odh Kserve Storage Initializer Rhel/Rhoai/Odh Kserve Agent Rhel/Rhoai/Odh Kserve Controller Rhel/Rhoai/Odh Kserve Router Rhel/Rhoai/Odh Feature Server Rhel/Rhoai/Odh Trustyai Nemo Guardrails Server Rhel/Rhaiis/Vllm Tpu Rhel/Rhoai/Odh Training Cuda128 Torch29 Py312 Rhel/Rhoai/Odh Training Rocm64 Torch29 Py312 Rhel/Rhaiis/Vllm Cpu Rhel/Rhaiis/Vllm Neuron Rhel/Ansible Automation Platform 25/Lightspeed Chatbot Rhel/Openshift Lightspeed/Lightspeed Ocp Rag Rhel/Rhoai/Odh Th06 Cpu Torch291 Py312 Rhel/Rhoai/Odh Th06 Cuda130 Torch291 Py312 Rhel/Rhoai/Odh Vllm Gaudi Rhel/Python3.12 Pillow/Rhaii/Vllm Gaudi Rhel/Rhaii/Vllm Neuron Rhel/Rhaii/Vllm Tpu Rhel/Lightspeed Core/Lightspeed Stack Rhel/Rhoai/Odh Kserve Autogluon Server Rhel/Rhai Early Access/Docling Sdk Cuda Rhel/Rhai Early Access/Docling Serve Cuda Rhel/Rhoai/Odh Llm D Kv Cache Rhel/Rhoai/Odh Th06 Cpu Torch210 Py312 Rhel/Rhoai/Odh Th06 Cuda130 Torch210 Py312 Rhel/Rhoai/Odh Th06 Rocm64 Torch291 Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Llmcompressor Cuda Py312 Rhel/Ask Sre/Rhel Cla/Rlsapi Rhel/Rhaiis/Model Opt Cuda Rhel/Rhaiis/Vllm Cuda Rhel/Rhaiis/Vllm Rocm Rhel/Rhaii/Model Opt Cuda Rhel/Rhaii/Vllm Cpu Rhel/Rhaii/Vllm Cuda Rhel/Rhaii/Vllm Rocm Rhel/Rhaii/Vllm Spyre Rhel/Ansible Automation Platform/Automation Dashboard Rhel/Ansible Automation Platform 26/Hub Rhel/Ansible Automation Platform 26/Lightspeed Chatbot Rhel/Ansible Automation Platform 27/Hub Rhel/Ansible Automation Platform 27/Lightspeed Chatbot Rhel/Galaxy Ng/Automation Hub Container/Aap Rag Content/Ansible Chatbot Stack/Koku Report Emailer/Exploit Intelligence/Vulnerability Analysis Rhel/Insights Host Inventory/Google Lightspeed Agent/Lscore Deploy/Lightspeed Core/Rag Tool Cpu Rhel/Lightspeed Core/Rag Tool Cuda 12.9 Rhel/Lightspeed Services Ocp/Ocp Rules Rhel/Openshift Lightspeed Tech Preview/Lightspeed Rag Tool Rhel/Openshift Lightspeed/Lightspeed Service Api Rhel/Rhoai/Odh Caikit Tgis Serving Rhel/Rhoai/Odh Llama Stack Core Rhel/Rhoai/Odh Pipeline Runtime Pytorch Llmcompressor Cuda Py312 Rhel/Rhoai/Odh Automl Rhel/Rhoai/Odh Autorag Rhel/Rhoai/Odh Mlflow Rhel/Rhoai/Odh Pipeline Runtime Datascience Cpu Py312 Rhel/Rhoai/Odh Pipeline Runtime Pytorch Cuda Py312 Rhel/Rhoai/Odh Pipeline Runtime Pytorch Rocm Py312 Rhel/Rhoai/Odh Pipeline Runtime Tensorflow Cuda Py312 Rhel/Rhoai/Odh Pipeline Runtime Tensorflow Rocm Py312 Rhel/Rhoai/Odh Trustyai Garak Lls Provider Dsp Rhel/Rhoai/Odh Workbench Codeserver Datascience Cpu Py312 Rhel/Rhoai/Odh Workbench Jupyter Datascience Cpu Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Cuda Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Rocm Py312 Rhel/Rhoai/Odh Workbench Jupyter Tensorflow Cuda Py312 Rhel/Rhoai/Odh Workbench Jupyter Tensorflow Rocm Py312 Rhel/Rhoai/Odh Workbench Jupyter Trustyai Cpu Py312 Rhel/Rhai/Guidellm Rhel/Rhoai/Odh Ogx Core Rhel/Rhoai/Odh Th Torch Cpu Py312 Rhel/Rhoai/Odh Th Torch Cuda Py312 Rhel/Rhoai/Odh Th Torch Rocm Py312 Rhel/Quay/Quay Service Tool/Quay/Quay Rhel/Rhelai3/Bootc Aws Cuda Rhel/Rhelai3/Bootc Azure Cuda Rhel/Rhelai3/Bootc Azure Rocm Rhel/Rhelai3/Bootc Gcp Cuda Rhel builds), which validate computed frame dimensions with _decompression_bomb_check before updating self._size, closing crafted GIF image with oversized frame extents. Until patched, this is a concern on applications parsing untrusted GIF images that catch DecompressionBombError exceptions and continue operating on the same image object, so avoid catching DecompressionBombError to reuse the same image instance, and ensure failed image processing operations immediately destroy or discard the affected object. As a compensating measure, monitor cgroup memory pressure via memory.events and PSI to detect anomalous memory spikes and pending OOM events, restrict process virtual address space using ulimit or prlimit (RLIMIT_AS) to bound oversized allocations, and enforce cgroups v2 memory resource limits on containers or process groups to prevent host-wide memory exhaustion. These controls limit blast radius and contain exploitation, but they do not fix the underlying retained image dimension defect — apply the upstream patch to fully remediate.\n\nHighest-impact CME: CME-1010 cgroup Memory Pressure Monitoring (memory.events / PSI) (medium)","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"scores":[{"cvss_v3":{"attackComplexity":"LOW","attackVector":"NETWORK","availabilityImpact":"HIGH","baseScore":7.5,"baseSeverity":"HIGH","confidentialityImpact":"NONE","integrityImpact":"NONE","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"NONE","vectorString":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Important"}],"title":"pillow: pillow: Denial of Service via memory exhaustion from crafted GIF image"},{"cve":"CVE-2026-112401","cwe":{"id":"CWE-770","name":"Allocation of Resources Without Limits or Throttling"},"discovery_date":"2026-10-19T21:21:13.210266+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2555861"}],"notes":[{"category":"description","text":"A flaw was found in Pillow. When parsing GIF images with numerous comment blocks, the library repeatedly copies data in an inefficient manner. A remote attacker can exploit this vulnerability by submitting a specially crafted GIF image to an application that processes images using Pillow, causing excessive CPU consumption and leading to a Denial of Service (DoS).","title":"Vulnerability description"},{"category":"summary","text":"pillow: pillow: Denial of Service via crafted GIF comments","title":"Vulnerability summary"},{"category":"other","text":"This vulnerability is rated Moderate severity because exploitation can exhaust worker CPU resources and trigger application unresponsiveness without compromising data confidentiality, integrity, or permitting arbitrary code execution. In Red Hat environments, Pillow is utilized across containerized workloads, machine learning pipelines, and automation platforms to process multimedia inputs. Because the algorithmic delay is restricted to the specific worker thread or container handling the untrusted image, broader platform stability is typically preserved by standard request timeouts and concurrency limits.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-112401"},{"category":"external","summary":"RHBZ#2555861","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2555861"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-112401","url":"https://www.cve.org/CVERecord?id=CVE-2026-112401"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-112401","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-112401"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/commit/68a3fe957d8a4d552e8c69d96066d749f0614c7e","url":"https://github.com/python-pillow/Pillow/commit/68a3fe957d8a4d552e8c69d96066d749f0614c7e"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0","url":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-74r2-7ff9-mjj2","url":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-74r2-7ff9-mjj2"}],"release_date":"2026-10-19T21:15:21.484000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"}],"scores":[{"cvss_v3":{"attackComplexity":"LOW","attackVector":"NETWORK","availabilityImpact":"LOW","baseScore":5.3,"baseSeverity":"MEDIUM","confidentialityImpact":"NONE","integrityImpact":"NONE","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"NONE","vectorString":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:L","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Moderate"}],"title":"pillow: pillow: Denial of Service via crafted GIF comments"},{"cve":"CVE-2026-112402","cwe":{"id":"CWE-1341","name":"Multiple Releases of Same Resource or Handle"},"discovery_date":"2026-10-19T21:21:01.037991+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2555859"}],"notes":[{"category":"description","text":"A flaw was found in Pillow. A remote attacker who can submit large image transformation operations under memory pressure can trigger improper error handling, leading to a double-free vulnerability (releasing the same memory block twice). This issue can result in an application crash, causing a Denial of Service (DoS), and may potentially lead to memory corruption.","title":"Vulnerability description"},{"category":"summary","text":"pillow: pillow: Denial of service via double free during image transformation","title":"Vulnerability summary"},{"category":"other","text":"This vulnerability is rated as Important because remote, unauthenticated attackers can cause service disruptions or memory corruption in applications that accept user-controlled image transformation requests without requiring prior privileges. In typical Red Hat environments, such as containerized AI or automation workloads processing image data, processing oversized transformation geometries under strict process memory limits forces premature deallocation of internal structures, culminating in process termination. Applications that sanitize image bounds or avoid exposing custom affine transform endpoints remain largely unaffected by this crash condition.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-112402"},{"category":"external","summary":"RHBZ#2555859","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2555859"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-112402","url":"https://www.cve.org/CVERecord?id=CVE-2026-112402"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-112402","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-112402"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/commit/aafa6c75f7cdabb9f409a8aac7c780bea31d08d9","url":"https://github.com/python-pillow/Pillow/commit/aafa6c75f7cdabb9f409a8aac7c780bea31d08d9"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0","url":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-7mwm-ghpf-9vrm","url":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-7mwm-ghpf-9vrm"}],"release_date":"2026-10-19T21:18:57.027000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"},{"category":"workaround","details":"--- CME Mitigation Guidance (2026-10-20) ---\nDisclaimer: The following mitigation solutions are provided as a courtesy, pending the application of an official patch. This content was generated utilizing AI tools in conjunction with data from https://cmetaxonomy.org. Customers are advised to independently evaluate and test these temporary mitigations within their own infrastructure.\n\nThe highest-value action is to eliminate premature destination image deletion in the ImagingScaleAffine error handling path so that the caller retains exclusive ownership of image memory cleanup: upgrade to Pillow 13.0.0 (or the fixed Python Pillow/Python3.11 Pillow/Rhoai/Odh Openvino Model Server Rhel/Rhelai3/Bootc Cuda Rhel/Rhelai3/Bootc Gaudi Rhel/Rhelai3/Bootc Rocm Rhel/Rhelai3/Disk Image Cuda Rhel/Python3X Pillow/Rhaiis/Vllm Spyre Rhel/Rhoai/Odh Kserve Storage Initializer Rhel/Rhoai/Odh Kserve Agent Rhel/Rhoai/Odh Kserve Controller Rhel/Rhoai/Odh Kserve Router Rhel/Rhoai/Odh Feature Server Rhel/Rhoai/Odh Trustyai Nemo Guardrails Server Rhel/Rhaiis/Vllm Tpu Rhel/Rhoai/Odh Training Cuda128 Torch29 Py312 Rhel/Rhoai/Odh Training Rocm64 Torch29 Py312 Rhel/Rhaiis/Vllm Cpu Rhel/Rhaiis/Vllm Neuron Rhel/Ansible Automation Platform 25/Lightspeed Chatbot Rhel/Openshift Lightspeed/Lightspeed Ocp Rag Rhel/Rhoai/Odh Th06 Cpu Torch291 Py312 Rhel/Rhoai/Odh Th06 Cuda130 Torch291 Py312 Rhel/Rhoai/Odh Vllm Gaudi Rhel/Python3.12 Pillow/Rhaii/Vllm Gaudi Rhel/Rhaii/Vllm Neuron Rhel/Rhaii/Vllm Tpu Rhel/Lightspeed Core/Lightspeed Stack Rhel/Rhoai/Odh Kserve Autogluon Server Rhel/Rhai Early Access/Docling Sdk Cuda Rhel/Rhai Early Access/Docling Serve Cuda Rhel/Rhoai/Odh Llm D Kv Cache Rhel/Rhoai/Odh Th06 Cpu Torch210 Py312 Rhel/Rhoai/Odh Th06 Cuda130 Torch210 Py312 Rhel/Rhoai/Odh Th06 Rocm64 Torch291 Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Llmcompressor Cuda Py312 Rhel/Ask Sre/Rhel Cla/Rlsapi Rhel/Rhaiis/Model Opt Cuda Rhel/Rhaiis/Vllm Cuda Rhel/Rhaiis/Vllm Rocm Rhel/Rhaii/Model Opt Cuda Rhel/Rhaii/Vllm Cpu Rhel/Rhaii/Vllm Cuda Rhel/Rhaii/Vllm Rocm Rhel/Rhaii/Vllm Spyre Rhel/Ansible Automation Platform/Automation Dashboard Rhel/Ansible Automation Platform 26/Hub Rhel/Ansible Automation Platform 26/Lightspeed Chatbot Rhel/Ansible Automation Platform 27/Hub Rhel/Ansible Automation Platform 27/Lightspeed Chatbot Rhel/Galaxy Ng/Automation Hub Container/Aap Rag Content/Ansible Chatbot Stack/Koku Report Emailer/Exploit Intelligence/Vulnerability Analysis Rhel/Insights Host Inventory/Google Lightspeed Agent/Lscore Deploy/Lightspeed Core/Rag Tool Cpu Rhel/Lightspeed Core/Rag Tool Cuda 12.9 Rhel/Lightspeed Services Ocp/Ocp Rules Rhel/Openshift Lightspeed Tech Preview/Lightspeed Rag Tool Rhel/Openshift Lightspeed/Lightspeed Service Api Rhel/Rhoai/Odh Caikit Tgis Serving Rhel/Rhoai/Odh Llama Stack Core Rhel/Rhoai/Odh Pipeline Runtime Pytorch Llmcompressor Cuda Py312 Rhel/Rhoai/Odh Automl Rhel/Rhoai/Odh Autorag Rhel/Rhoai/Odh Mlflow Rhel/Rhoai/Odh Pipeline Runtime Datascience Cpu Py312 Rhel/Rhoai/Odh Pipeline Runtime Pytorch Cuda Py312 Rhel/Rhoai/Odh Pipeline Runtime Pytorch Rocm Py312 Rhel/Rhoai/Odh Pipeline Runtime Tensorflow Cuda Py312 Rhel/Rhoai/Odh Pipeline Runtime Tensorflow Rocm Py312 Rhel/Rhoai/Odh Trustyai Garak Lls Provider Dsp Rhel/Rhoai/Odh Workbench Codeserver Datascience Cpu Py312 Rhel/Rhoai/Odh Workbench Jupyter Datascience Cpu Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Cuda Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Rocm Py312 Rhel/Rhoai/Odh Workbench Jupyter Tensorflow Cuda Py312 Rhel/Rhoai/Odh Workbench Jupyter Tensorflow Rocm Py312 Rhel/Rhoai/Odh Workbench Jupyter Trustyai Cpu Py312 Rhel/Rhai/Guidellm Rhel/Rhoai/Odh Ogx Core Rhel/Rhoai/Odh Th Torch Cpu Py312 Rhel/Rhoai/Odh Th Torch Cuda Py312 Rhel/Rhoai/Odh Th Torch Rocm Py312 Rhel/Quay/Quay Service Tool/Quay/Quay Rhel/Rhelai3/Bootc Aws Cuda Rhel/Rhelai3/Bootc Azure Cuda Rhel/Rhelai3/Bootc Azure Rocm Rhel/Rhelai3/Bootc Gcp Cuda Rhel builds), which remove the ImagingDelete(imOut) call from the allocation error path in ImagingScaleAffine() and rely on the caller to free the destination image when the transform fails, closing all 2 injection vectors (crafted Image.transform() operations using Transform.EXTENT, crafted Image.transform() operations using Transform.AFFINE). Until patched, this is a concern on Applications executing Pillow Image.transform() operations with untrusted dimensions under memory pressure, so Validate and constrain requested dimensions in Image.transform() calls, avoid using Transform.EXTENT or scale-and-translate Transform.AFFINE with NEAREST resampling on untrusted input, and ensure sufficient memory limits to prevent selective allocation failures. As a compensating measure, Deploy EDR agents to detect abnormal process terminations and memory corruption exploitation attempts resulting from the double-free condition. These controls limit blast radius and contain exploitation, but they do not fix the underlying double-free error handling defect — apply the upstream patch to fully remediate.\n\nHighest-impact CME: CME-1001 EDR Agent (Endpoint Detection & Response) (medium)","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"scores":[{"cvss_v3":{"attackComplexity":"LOW","attackVector":"NETWORK","availabilityImpact":"HIGH","baseScore":7.5,"baseSeverity":"HIGH","confidentialityImpact":"NONE","integrityImpact":"NONE","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"NONE","vectorString":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Important"}],"title":"pillow: pillow: Denial of service via double free during image transformation"},{"cve":"CVE-2026-112403","cwe":{"id":"CWE-787","name":"Out-of-bounds Write"},"discovery_date":"2026-10-19T21:41:54.085786+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2555871"}],"notes":[{"category":"description","text":"A flaw was found in Pillow. A remote attacker could exploit this vulnerability to achieve arbitrary code execution or cause a Denial of Service (DoS) by supplying a specially crafted lossless JPEG image to an application that downscales images or generates thumbnails. This issue occurs because the JPEG decoder fails to verify actual output dimensions against the allocated memory buffer during scaling, causing full-width image data to be written beyond buffer boundaries. Successful exploitation leads to heap memory corruption, resulting in an application crash or potential code execution.","title":"Vulnerability description"},{"category":"summary","text":"pillow: Pillow: Heap buffer overflow in ImagingJpegDecode","title":"Vulnerability summary"},{"category":"other","text":"The overflow only triggers when the application calls `draft()` or `thumbnail()` before loading the image. Direct `Image.open().load()` or `Image.open().resize()` decode at scale 1 and are unaffected. Many image processing pipelines use `load()` followed by `resize()`, not `thumbnail()`. The libjpeg-turbo 3.x requirement, specific code path, and dimension threshold make this **not easily exploitable** across the entire Pillow install base.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-112403"},{"category":"external","summary":"RHBZ#2555871","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2555871"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-112403","url":"https://www.cve.org/CVERecord?id=CVE-2026-112403"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-112403","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-112403"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/commit/242a059aedc1935097d9aa1dc3f099750467ba75","url":"https://github.com/python-pillow/Pillow/commit/242a059aedc1935097d9aa1dc3f099750467ba75"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0","url":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-8g5g-8vmr-9782","url":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-8g5g-8vmr-9782"}],"release_date":"2026-10-19T21:22:21.077000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"}],"scores":[{"cvss_v3":{"attackComplexity":"HIGH","attackVector":"NETWORK","availabilityImpact":"HIGH","baseScore":8.1,"baseSeverity":"HIGH","confidentialityImpact":"HIGH","integrityImpact":"HIGH","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"NONE","vectorString":"CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:H/I:H/A:H","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Important"}],"title":"pillow: Pillow: Heap buffer overflow in ImagingJpegDecode"},{"cve":"CVE-2026-112404","cwe":{"id":"CWE-125","name":"Out-of-bounds Read"},"discovery_date":"2026-10-19T21:43:45.124371+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2555874"}],"notes":[{"category":"description","text":"A flaw was found in Pillow. When processing Tagged Image File Format (TIFF) images, the library fails to properly validate buffer boundaries during tile and color decoding routines. A remote attacker could exploit this vulnerability by supplying a specially crafted TIFF file with conflicting metadata tags, causing an out-of-bounds read. Successful exploitation can lead to a Denial of Service (DoS) by crashing the application, or result in information disclosure by exposing adjacent heap memory in the output image.","title":"Vulnerability description"},{"category":"summary","text":"pillow: Pillow: Denial of Service via out-of-bounds read in TIFF decoding","title":"Vulnerability summary"},{"category":"other","text":"This vulnerability is rated as an Important severity flaw because an unauthenticated remote attacker can exploit it to terminate backend worker processes or leak adjacent heap contents without requiring elevated privileges. Red Hat products utilizing Pillow within containerized services, AI model serving, or automation workflows may be exposed if they accept and parse untrusted TIFF files. Because TIFF parsing occurs automatically during standard image opening routines without requiring custom decode flags, services that ingest user-provided media are susceptible to denial of service or memory exposure.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-112404"},{"category":"external","summary":"RHBZ#2555874","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2555874"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-112404","url":"https://www.cve.org/CVERecord?id=CVE-2026-112404"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-112404","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-112404"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/commit/1d4773e6cea0c4a2814e6d4cdc099bbcf56da6ce","url":"https://github.com/python-pillow/Pillow/commit/1d4773e6cea0c4a2814e6d4cdc099bbcf56da6ce"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0","url":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-c6hw-5grf-ff96","url":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-c6hw-5grf-ff96"}],"release_date":"2026-10-19T21:27:56.705000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"},{"category":"workaround","details":"--- CME Mitigation Guidance (2026-10-20) ---\nDisclaimer: The following mitigation solutions are provided as a courtesy, pending the application of an official patch. This content was generated utilizing AI tools in conjunction with data from https://cmetaxonomy.org. Customers are advised to independently evaluate and test these temporary mitigations within their own infrastructure.\n\nThe highest-value action is to block untrusted TIFF ingestion or ensure row size validation between unpackers and libtiff buffers in Pillow decoding workflows: upgrade to Pillow 13.0.0 (or the fixed Python Pillow/Python3.11 Pillow/Rhoai/Odh Openvino Model Server Rhel/Rhelai3/Bootc Cuda Rhel/Rhelai3/Bootc Gaudi Rhel/Rhelai3/Bootc Rocm Rhel/Rhelai3/Disk Image Cuda Rhel/Python3X Pillow/Rhaiis/Vllm Spyre Rhel/Rhoai/Odh Kserve Storage Initializer Rhel/Rhoai/Odh Kserve Agent Rhel/Rhoai/Odh Kserve Controller Rhel/Rhoai/Odh Kserve Router Rhel/Rhoai/Odh Feature Server Rhel/Rhoai/Odh Trustyai Nemo Guardrails Server Rhel/Rhaiis/Vllm Tpu Rhel/Rhoai/Odh Training Cuda128 Torch29 Py312 Rhel/Rhoai/Odh Training Rocm64 Torch29 Py312 Rhel/Rhaiis/Vllm Cpu Rhel/Rhaiis/Vllm Neuron Rhel/Ansible Automation Platform 25/Lightspeed Chatbot Rhel/Openshift Lightspeed/Lightspeed Ocp Rag Rhel/Rhoai/Odh Th06 Cpu Torch291 Py312 Rhel/Rhoai/Odh Th06 Cuda130 Torch291 Py312 Rhel/Rhoai/Odh Vllm Gaudi Rhel/Python3.12 Pillow/Rhaii/Vllm Gaudi Rhel/Rhaii/Vllm Neuron Rhel/Rhaii/Vllm Tpu Rhel/Lightspeed Core/Lightspeed Stack Rhel/Rhoai/Odh Kserve Autogluon Server Rhel/Rhai Early Access/Docling Sdk Cuda Rhel/Rhai Early Access/Docling Serve Cuda Rhel/Rhoai/Odh Llm D Kv Cache Rhel/Rhoai/Odh Th06 Cpu Torch210 Py312 Rhel/Rhoai/Odh Th06 Cuda130 Torch210 Py312 Rhel/Rhoai/Odh Th06 Rocm64 Torch291 Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Llmcompressor Cuda Py312 Rhel/Ask Sre/Rhel Cla/Rlsapi Rhel/Rhaiis/Model Opt Cuda Rhel/Rhaiis/Vllm Cuda Rhel/Rhaiis/Vllm Rocm Rhel/Rhaii/Model Opt Cuda Rhel/Rhaii/Vllm Cpu Rhel/Rhaii/Vllm Cuda Rhel/Rhaii/Vllm Rocm Rhel/Rhaii/Vllm Spyre Rhel/Ansible Automation Platform/Automation Dashboard Rhel/Ansible Automation Platform 26/Hub Rhel/Ansible Automation Platform 26/Lightspeed Chatbot Rhel/Ansible Automation Platform 27/Hub Rhel/Ansible Automation Platform 27/Lightspeed Chatbot Rhel/Galaxy Ng/Automation Hub Container/Aap Rag Content/Ansible Chatbot Stack/Koku Report Emailer/Exploit Intelligence/Vulnerability Analysis Rhel/Insights Host Inventory/Google Lightspeed Agent/Lscore Deploy/Lightspeed Core/Rag Tool Cpu Rhel/Lightspeed Core/Rag Tool Cuda 12.9 Rhel/Lightspeed Services Ocp/Ocp Rules Rhel/Openshift Lightspeed Tech Preview/Lightspeed Rag Tool Rhel/Openshift Lightspeed/Lightspeed Service Api Rhel/Rhoai/Odh Caikit Tgis Serving Rhel/Rhoai/Odh Llama Stack Core Rhel/Rhoai/Odh Pipeline Runtime Pytorch Llmcompressor Cuda Py312 Rhel/Rhoai/Odh Automl Rhel/Rhoai/Odh Autorag Rhel/Rhoai/Odh Mlflow Rhel/Rhoai/Odh Pipeline Runtime Datascience Cpu Py312 Rhel/Rhoai/Odh Pipeline Runtime Pytorch Cuda Py312 Rhel/Rhoai/Odh Pipeline Runtime Pytorch Rocm Py312 Rhel/Rhoai/Odh Pipeline Runtime Tensorflow Cuda Py312 Rhel/Rhoai/Odh Pipeline Runtime Tensorflow Rocm Py312 Rhel/Rhoai/Odh Trustyai Garak Lls Provider Dsp Rhel/Rhoai/Odh Workbench Codeserver Datascience Cpu Py312 Rhel/Rhoai/Odh Workbench Jupyter Datascience Cpu Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Cuda Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Rocm Py312 Rhel/Rhoai/Odh Workbench Jupyter Tensorflow Cuda Py312 Rhel/Rhoai/Odh Workbench Jupyter Tensorflow Rocm Py312 Rhel/Rhoai/Odh Workbench Jupyter Trustyai Cpu Py312 Rhel/Rhai/Guidellm Rhel/Rhoai/Odh Ogx Core Rhel/Rhoai/Odh Th Torch Cpu Py312 Rhel/Rhoai/Odh Th Torch Cuda Py312 Rhel/Rhoai/Odh Th Torch Rocm Py312 Rhel/Quay/Quay Service Tool/Quay/Quay Rhel/Rhelai3/Bootc Aws Cuda Rhel/Rhelai3/Bootc Azure Cuda Rhel/Rhelai3/Bootc Azure Rocm Rhel/Rhelai3/Bootc Gcp Cuda Rhel builds), which reject TIFF images where the unpacker row size exceeds libtiff's decoded row and zero unwritten buffer space in tile and strip decoders, closing all 2 injection vectors (crafted TIFF image file with duplicate PhotometricInterpretation tags, crafted TIFF image file with duplicate BitsPerSample tags). Until patched, this is a concern on applications and automated pipelines parsing untrusted TIFF images with Pillow, so restrict untrusted TIFF file processing and disable automated thumbnail or preview generation for untrusted TIFF inputs. These controls limit blast radius and contain exploitation, but they do not fix the underlying heap out-of-bounds read defect — apply the upstream patch to fully remediate.","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"scores":[{"cvss_v3":{"attackComplexity":"LOW","attackVector":"NETWORK","availabilityImpact":"HIGH","baseScore":8.2,"baseSeverity":"HIGH","confidentialityImpact":"LOW","integrityImpact":"NONE","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"NONE","vectorString":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:N/A:H","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Important"}],"title":"pillow: Pillow: Denial of Service via out-of-bounds read in TIFF decoding"},{"cve":"CVE-2026-112405","cwe":{"id":"CWE-476","name":"NULL Pointer Dereference"},"discovery_date":"2026-10-19T21:42:15.131597+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2555873"}],"notes":[{"category":"description","text":"A flaw was found in Pillow. A remote attacker can cause a Denial of Service (DoS) by supplying a specially crafted image file to an application that processes images with the library. This issue occurs because the Image.merge() function fails to properly validate the first input band, allowing mismatched image storage formats to trigger a null pointer dereference. Consequently, processing the malformed image causes the native application process to crash.","title":"Vulnerability description"},{"category":"summary","text":"pillow: Pillow: Denial of Service via null pointer dereference in Image.merge()","title":"Vulnerability summary"},{"category":"other","text":"This vulnerability is rated as having an Important impact for Red Hat products because an unauthenticated remote attacker can crash service worker processes handling user uploads without requiring privileges or prior interaction. In Red Hat environments such as Ansible Automation Platform, Satellite, and OpenShift AI, Pillow is embedded across containerized services and platform tooling that handle image ingest and transformation. Because parsing an invalid image file triggers an immediate segmentation fault within the native extension, the issue causes service disruption, although it does not compromise data confidentiality or allow arbitrary code execution.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-112405"},{"category":"external","summary":"RHBZ#2555873","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2555873"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-112405","url":"https://www.cve.org/CVERecord?id=CVE-2026-112405"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-112405","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-112405"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/commit/b6ee66e8b2c2c0932bdf65f963268d9f2927ef86","url":"https://github.com/python-pillow/Pillow/commit/b6ee66e8b2c2c0932bdf65f963268d9f2927ef86"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0","url":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-cp9v-662q-vvx9","url":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-cp9v-662q-vvx9"}],"release_date":"2026-10-19T21:29:53.515000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"},{"category":"workaround","details":"--- CME Mitigation Guidance (2026-10-20) ---\nDisclaimer: The following mitigation solutions are provided as a courtesy, pending the application of an official patch. This content was generated utilizing AI tools in conjunction with data from https://cmetaxonomy.org. Customers are advised to independently evaluate and test these temporary mitigations within their own infrastructure.\n\nThe highest-value action is to validate all input image bands to reject 4-byte pixel sizes and 32-bit storage modes prior to invoking Image.merge(): upgrade to Pillow 13.0.0 (or the fixed Python Pillow/Python3.11 Pillow/Rhoai/Odh Openvino Model Server Rhel/Rhelai3/Bootc Cuda Rhel/Rhelai3/Bootc Gaudi Rhel/Rhelai3/Bootc Rocm Rhel/Rhelai3/Disk Image Cuda Rhel/Python3X Pillow/Rhaiis/Vllm Spyre Rhel/Rhoai/Odh Kserve Storage Initializer Rhel/Rhoai/Odh Kserve Agent Rhel/Rhoai/Odh Kserve Controller Rhel/Rhoai/Odh Kserve Router Rhel/Rhoai/Odh Feature Server Rhel/Rhoai/Odh Trustyai Nemo Guardrails Server Rhel/Rhaiis/Vllm Tpu Rhel/Rhoai/Odh Training Cuda128 Torch29 Py312 Rhel/Rhoai/Odh Training Rocm64 Torch29 Py312 Rhel/Rhaiis/Vllm Cpu Rhel/Rhaiis/Vllm Neuron Rhel/Ansible Automation Platform 25/Lightspeed Chatbot Rhel/Openshift Lightspeed/Lightspeed Ocp Rag Rhel/Rhoai/Odh Th06 Cpu Torch291 Py312 Rhel/Rhoai/Odh Th06 Cuda130 Torch291 Py312 Rhel/Rhoai/Odh Vllm Gaudi Rhel/Python3.12 Pillow/Rhaii/Vllm Gaudi Rhel/Rhaii/Vllm Neuron Rhel/Rhaii/Vllm Tpu Rhel/Lightspeed Core/Lightspeed Stack Rhel/Rhoai/Odh Kserve Autogluon Server Rhel/Rhai Early Access/Docling Sdk Cuda Rhel/Rhai Early Access/Docling Serve Cuda Rhel/Rhoai/Odh Llm D Kv Cache Rhel/Rhoai/Odh Th06 Cpu Torch210 Py312 Rhel/Rhoai/Odh Th06 Cuda130 Torch210 Py312 Rhel/Rhoai/Odh Th06 Rocm64 Torch291 Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Llmcompressor Cuda Py312 Rhel/Ask Sre/Rhel Cla/Rlsapi Rhel/Rhaiis/Model Opt Cuda Rhel/Rhaiis/Vllm Cuda Rhel/Rhaiis/Vllm Rocm Rhel/Rhaii/Model Opt Cuda Rhel/Rhaii/Vllm Cpu Rhel/Rhaii/Vllm Cuda Rhel/Rhaii/Vllm Rocm Rhel/Rhaii/Vllm Spyre Rhel/Ansible Automation Platform/Automation Dashboard Rhel/Ansible Automation Platform 26/Hub Rhel/Ansible Automation Platform 26/Lightspeed Chatbot Rhel/Ansible Automation Platform 27/Hub Rhel/Ansible Automation Platform 27/Lightspeed Chatbot Rhel/Galaxy Ng/Automation Hub Container/Aap Rag Content/Ansible Chatbot Stack/Koku Report Emailer/Exploit Intelligence/Vulnerability Analysis Rhel/Insights Host Inventory/Google Lightspeed Agent/Lscore Deploy/Lightspeed Core/Rag Tool Cpu Rhel/Lightspeed Core/Rag Tool Cuda 12.9 Rhel/Lightspeed Services Ocp/Ocp Rules Rhel/Openshift Lightspeed Tech Preview/Lightspeed Rag Tool Rhel/Openshift Lightspeed/Lightspeed Service Api Rhel/Rhoai/Odh Caikit Tgis Serving Rhel/Rhoai/Odh Llama Stack Core Rhel/Rhoai/Odh Pipeline Runtime Pytorch Llmcompressor Cuda Py312 Rhel/Rhoai/Odh Automl Rhel/Rhoai/Odh Autorag Rhel/Rhoai/Odh Mlflow Rhel/Rhoai/Odh Pipeline Runtime Datascience Cpu Py312 Rhel/Rhoai/Odh Pipeline Runtime Pytorch Cuda Py312 Rhel/Rhoai/Odh Pipeline Runtime Pytorch Rocm Py312 Rhel/Rhoai/Odh Pipeline Runtime Tensorflow Cuda Py312 Rhel/Rhoai/Odh Pipeline Runtime Tensorflow Rocm Py312 Rhel/Rhoai/Odh Trustyai Garak Lls Provider Dsp Rhel/Rhoai/Odh Workbench Codeserver Datascience Cpu Py312 Rhel/Rhoai/Odh Workbench Jupyter Datascience Cpu Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Cuda Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Rocm Py312 Rhel/Rhoai/Odh Workbench Jupyter Tensorflow Cuda Py312 Rhel/Rhoai/Odh Workbench Jupyter Tensorflow Rocm Py312 Rhel/Rhoai/Odh Workbench Jupyter Trustyai Cpu Py312 Rhel/Rhai/Guidellm Rhel/Rhoai/Odh Ogx Core Rhel/Rhoai/Odh Th Torch Cpu Py312 Rhel/Rhoai/Odh Th Torch Cuda Py312 Rhel/Rhoai/Odh Th Torch Rocm Py312 Rhel/Quay/Quay Service Tool/Quay/Quay Rhel/Rhelai3/Bootc Aws Cuda Rhel/Rhelai3/Bootc Azure Cuda Rhel/Rhelai3/Bootc Azure Rocm Rhel/Rhelai3/Bootc Gcp Cuda Rhel builds), which validate all input bands and reject any band with a 4-byte pixel size or non-single band configuration with a mode error, closing all 2 injection vectors (crafted IPTC image file parsed during standard open and load operations, direct invocation of Image.merge() with a 32-bit image band). Until patched, this is a concern on applications using Pillow to process untrusted image files or calling Image.merge() on untrusted image bands, so restrict image ingestion pipelines from processing untrusted IPTC files or passing unvalidated 32-bit single-band images into Image.merge(). As a compensating measure, enforce an explicit allowlist of permitted image modes and band configurations before invoking Image.merge() to prevent processing unsupported 32-bit image bands. These controls limit blast radius and contain exploitation, but they do not fix the underlying null pointer dereference defect — apply the upstream patch to fully remediate.\n\nHighest-impact CME: CME-1315 Allowlist-Preferred Input Validation (Deny-by-Default Filtering) (medium)","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"scores":[{"cvss_v3":{"attackComplexity":"LOW","attackVector":"NETWORK","availabilityImpact":"HIGH","baseScore":7.5,"baseSeverity":"HIGH","confidentialityImpact":"NONE","integrityImpact":"NONE","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"NONE","vectorString":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Important"}],"title":"pillow: Pillow: Denial of Service via null pointer dereference in Image.merge()"},{"cve":"CVE-2026-112406","cwe":{"id":"CWE-835","name":"Loop with Unreachable Exit Condition"},"discovery_date":"2026-10-20T16:21:58.830847+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2556322"}],"notes":[{"category":"description","text":"A flaw was found in Pillow. This vulnerability allows a remote attacker to cause a Denial of Service (DoS) through an infinite loop during PDF page processing. When rewriting pages, the parser fails to detect circular references in the page tree hierarchy. By submitting a specially crafted PDF document to an application that modifies or appends images to it, an attacker can exhaust system CPU resources and make the service unresponsive.","title":"Vulnerability description"},{"category":"summary","text":"pillow: pillow: Denial of Service via infinite loop during PDF page rewriting","title":"Vulnerability summary"},{"category":"other","text":"This vulnerability is rated as Important because an unauthenticated remote user can supply a crafted document to trigger an infinite processing loop, resulting in complete CPU core exhaustion and denial of service for the target application worker. The issue specifically threatens services or workloads that ingest untrusted PDFs and append or save pages to them. Standard image parsing and reading pipelines that do not modify or append to existing PDF structures remain unaffected.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-112406"},{"category":"external","summary":"RHBZ#2556322","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2556322"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-112406","url":"https://www.cve.org/CVERecord?id=CVE-2026-112406"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-112406","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-112406"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/commit/b3ebf20c694f2f12537a629c30229a1d5a53f777","url":"https://github.com/python-pillow/Pillow/commit/b3ebf20c694f2f12537a629c30229a1d5a53f777"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0","url":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-f2g6-542v-9j58","url":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-f2g6-542v-9j58"}],"release_date":"2026-10-19T21:31:49.946000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"},{"category":"workaround","details":"--- CME Mitigation Guidance (2026-10-20) ---\nDisclaimer: The following mitigation solutions are provided as a courtesy, pending the application of an official patch. This content was generated utilizing AI tools in conjunction with data from https://cmetaxonomy.org. Customers are advised to independently evaluate and test these temporary mitigations within their own infrastructure.\n\nThe highest-value action is to halt page-tree traversal upon detecting repeated node references during PDF page rewriting in Pillow: upgrade to Pillow 13.0.0 (or the fixed Python Pillow/Python3.11 Pillow/Rhoai/Odh Openvino Model Server Rhel/Rhelai3/Bootc Cuda Rhel/Rhelai3/Bootc Gaudi Rhel/Rhelai3/Bootc Rocm Rhel/Rhelai3/Disk Image Cuda Rhel/Python3X Pillow/Rhaiis/Vllm Spyre Rhel/Rhoai/Odh Kserve Storage Initializer Rhel/Rhoai/Odh Kserve Agent Rhel/Rhoai/Odh Kserve Controller Rhel/Rhoai/Odh Kserve Router Rhel/Rhoai/Odh Feature Server Rhel/Rhoai/Odh Trustyai Nemo Guardrails Server Rhel/Rhaiis/Vllm Tpu Rhel/Rhoai/Odh Training Cuda128 Torch29 Py312 Rhel/Rhoai/Odh Training Rocm64 Torch29 Py312 Rhel/Rhaiis/Vllm Cpu Rhel/Rhaiis/Vllm Neuron Rhel/Ansible Automation Platform 25/Lightspeed Chatbot Rhel/Openshift Lightspeed/Lightspeed Ocp Rag Rhel/Rhoai/Odh Th06 Cpu Torch291 Py312 Rhel/Rhoai/Odh Th06 Cuda130 Torch291 Py312 Rhel/Rhoai/Odh Vllm Gaudi Rhel/Python3.12 Pillow/Rhaii/Vllm Gaudi Rhel/Rhaii/Vllm Neuron Rhel/Rhaii/Vllm Tpu Rhel/Lightspeed Core/Lightspeed Stack Rhel/Rhoai/Odh Kserve Autogluon Server Rhel/Rhai Early Access/Docling Sdk Cuda Rhel/Rhai Early Access/Docling Serve Cuda Rhel/Rhoai/Odh Llm D Kv Cache Rhel/Rhoai/Odh Th06 Cpu Torch210 Py312 Rhel/Rhoai/Odh Th06 Cuda130 Torch210 Py312 Rhel/Rhoai/Odh Th06 Rocm64 Torch291 Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Llmcompressor Cuda Py312 Rhel/Ask Sre/Rhel Cla/Rlsapi Rhel/Rhaiis/Model Opt Cuda Rhel/Rhaiis/Vllm Cuda Rhel/Rhaiis/Vllm Rocm Rhel/Rhaii/Model Opt Cuda Rhel/Rhaii/Vllm Cpu Rhel/Rhaii/Vllm Cuda Rhel/Rhaii/Vllm Rocm Rhel/Rhaii/Vllm Spyre Rhel/Ansible Automation Platform/Automation Dashboard Rhel/Ansible Automation Platform 26/Hub Rhel/Ansible Automation Platform 26/Lightspeed Chatbot Rhel/Ansible Automation Platform 27/Hub Rhel/Ansible Automation Platform 27/Lightspeed Chatbot Rhel/Galaxy Ng/Automation Hub Container/Aap Rag Content/Ansible Chatbot Stack/Koku Report Emailer/Exploit Intelligence/Vulnerability Analysis Rhel/Insights Host Inventory/Google Lightspeed Agent/Lscore Deploy/Lightspeed Core/Rag Tool Cpu Rhel/Lightspeed Core/Rag Tool Cuda 12.9 Rhel/Lightspeed Services Ocp/Ocp Rules Rhel/Openshift Lightspeed Tech Preview/Lightspeed Rag Tool Rhel/Openshift Lightspeed/Lightspeed Service Api Rhel/Rhoai/Odh Caikit Tgis Serving Rhel/Rhoai/Odh Llama Stack Core Rhel/Rhoai/Odh Pipeline Runtime Pytorch Llmcompressor Cuda Py312 Rhel/Rhoai/Odh Automl Rhel/Rhoai/Odh Autorag Rhel/Rhoai/Odh Mlflow Rhel/Rhoai/Odh Trustyai Garak Lls Provider Dsp Rhel/Rhoai/Odh Workbench Codeserver Datascience Cpu Py312 Rhel/Rhoai/Odh Workbench Jupyter Datascience Cpu Py312 Rhel/Rhai/Guidellm Rhel/Rhoai/Odh Ogx Core Rhel/Rhoai/Odh Pipeline Runtime Datascience Cpu Py312 Rhel/Rhoai/Odh Pipeline Runtime Pytorch Cuda Py312 Rhel/Rhoai/Odh Pipeline Runtime Pytorch Rocm Py312 Rhel/Rhoai/Odh Pipeline Runtime Tensorflow Cuda Py312 Rhel/Rhoai/Odh Pipeline Runtime Tensorflow Rocm Py312 Rhel/Rhoai/Odh Th Torch Cpu Py312 Rhel/Rhoai/Odh Th Torch Cuda Py312 Rhel/Rhoai/Odh Th Torch Rocm Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Cuda Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Rocm Py312 Rhel/Rhoai/Odh Workbench Jupyter Tensorflow Cuda Py312 Rhel/Rhoai/Odh Workbench Jupyter Tensorflow Rocm Py312 Rhel/Rhoai/Odh Workbench Jupyter Trustyai Cpu Py312 Rhel/Quay/Quay Service Tool/Quay/Quay Rhel/Rhelai3/Bootc Aws Cuda Rhel/Rhelai3/Bootc Azure Cuda Rhel/Rhelai3/Bootc Azure Rocm Rhel/Rhelai3/Bootc Gcp Cuda Rhel builds), which track visited page-tree node references during deletion traversal and halt iteration if a reference repeats, closing crafted PDF with cyclic Parent references in page tree. Until patched, this is a concern on applications that modify or append images to supplied PDFs using Pillow, so restrict untrusted PDF submissions to applications that modify or append images, and apply execution timeouts on PDF handling workers. As a compensating measure, impose maximum CPU time limits on worker processes using RLIMIT_CPU via ulimit or prlimit to terminate execution when an infinite loop occurs, detect unresponsiveness in hung worker processes via systemd WatchdogSec heartbeats and automatically restart the service, and constrain maximum CPU quotas for the processing container or service using cgroups v2 to prevent host resource starvation. These controls limit blast radius and contain exploitation, but they do not fix the underlying infinite loop defect — apply the upstream patch to fully remediate.\n\nHighest-impact CME: CME-114 Process Resource Limits (ulimit/prlimit) (medium)","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"scores":[{"cvss_v3":{"attackComplexity":"LOW","attackVector":"NETWORK","availabilityImpact":"HIGH","baseScore":7.5,"baseSeverity":"HIGH","confidentialityImpact":"NONE","integrityImpact":"NONE","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"NONE","vectorString":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Important"}],"title":"pillow: pillow: Denial of Service via infinite loop during PDF page rewriting"},{"cve":"CVE-2026-112891","cwe":{"id":"CWE-770","name":"Allocation of Resources Without Limits or Throttling"},"discovery_date":"2026-10-20T14:01:03.148614+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2556193"}],"notes":[{"category":"description","text":"A flaw was found in Pillow. A remote attacker can cause a Denial of Service (DoS) by providing a specially crafted image with an excessively large comment to an application that saves it as a GIF. Due to inefficient data copying while assembling image headers, processing such comments consumes excessive CPU resources and can render the application unresponsive.","title":"Vulnerability description"},{"category":"summary","text":"pillow: pillow: Denial of Service via large GIF comments","title":"Vulnerability summary"},{"category":"other","text":"This vulnerability is assessed as having a Moderate impact because it results solely in a denial of service via CPU exhaustion without compromising system confidentiality or data integrity. In Red Hat environments, the affected library is utilized across various service and container workloads, but triggering the flaw requires an application to ingest untrusted image input and subsequently write or re-encode it using the GIF format. Workloads that do not accept external input or do not serialize images to GIF remain unaffected.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-112891"},{"category":"external","summary":"RHBZ#2556193","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2556193"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-112891","url":"https://www.cve.org/CVERecord?id=CVE-2026-112891"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-112891","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-112891"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/commit/68a3fe957d8a4d552e8c69d96066d749f0614c7e","url":"https://github.com/python-pillow/Pillow/commit/68a3fe957d8a4d552e8c69d96066d749f0614c7e"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0","url":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-fc9g-5qg7-mf45","url":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-fc9g-5qg7-mf45"}],"release_date":"2026-10-19T21:33:32.180000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"}],"scores":[{"cvss_v3":{"attackComplexity":"LOW","attackVector":"NETWORK","availabilityImpact":"LOW","baseScore":5.3,"baseSeverity":"MEDIUM","confidentialityImpact":"NONE","integrityImpact":"NONE","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"NONE","vectorString":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:L","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Moderate"}],"title":"pillow: pillow: Denial of Service via large GIF comments"},{"cve":"CVE-2026-112892","cwe":{"id":"CWE-409","name":"Improper Handling of Highly Compressed Data (Data Amplification)"},"discovery_date":"2026-10-22T21:32:36.389983+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2559393"}],"notes":[{"category":"description","text":"A flaw was found in Pillow. A remote attacker can cause a Denial of Service (DoS) by supplying a specially crafted Microsoft Paint (MSP) image file to an application that processes image pixels. Due to missing expansion limits during decompression, the run-length encoding (RLE) decoder expands image data into an unbounded memory buffer, resulting in memory exhaustion and application crashes.","title":"Vulnerability description"},{"category":"summary","text":"pillow: pillow: Denial of Service via unbounded memory expansion in MSP image decoder","title":"Vulnerability summary"},{"category":"other","text":"This vulnerability is evaluated as an Important impact because remote unauthenticated actors can induce an application denial of service by consuming all host or container memory, bypassing standard decompression safety limits. In Red Hat deployments such as Red Hat OpenShift AI and Red Hat Quay, service ingestion workflows accepting user-submitted graphics could suffer process termination or node instability when decoding untrusted assets. Deployments that process only verified internal content or do not expose image analysis to external inputs face substantially lower operational risk.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-112892"},{"category":"external","summary":"RHBZ#2559393","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2559393"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-112892","url":"https://www.cve.org/CVERecord?id=CVE-2026-112892"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-112892","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-112892"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/commit/5dcb630b9798efd91929cb873d677aeee6ed0bb3","url":"https://github.com/python-pillow/Pillow/commit/5dcb630b9798efd91929cb873d677aeee6ed0bb3"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0","url":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-gg2r-wvfm-q9rx","url":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-gg2r-wvfm-q9rx"}],"release_date":"2026-10-19T21:35:25.797000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"},{"category":"workaround","details":"Mitigation for this issue is either not available or the currently available options do not meet the Red Hat Product Security criteria comprising ease of use and deployment, applicability to widespread installation base, or stability.","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"scores":[{"cvss_v3":{"attackComplexity":"LOW","attackVector":"NETWORK","availabilityImpact":"HIGH","baseScore":7.5,"baseSeverity":"HIGH","confidentialityImpact":"NONE","integrityImpact":"NONE","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"NONE","vectorString":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Important"}],"title":"pillow: pillow: Denial of Service via unbounded memory expansion in MSP image decoder"},{"cve":"CVE-2026-112893","cwe":{"id":"CWE-763","name":"Release of Invalid Pointer or Reference"},"discovery_date":"2026-10-19T22:01:14.071566+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2555878"}],"notes":[{"category":"description","text":"A flaw was found in Pillow. When rendering text using FreeType with an active stroke width, a failure during glyph loading can cause the library to free an uninitialized or stale memory pointer. A remote attacker could exploit this vulnerability by supplying a malformed font to an application that renders stroked text. Successful exploitation can lead to memory corruption or an application crash, resulting in a Denial of Service (DoS).","title":"Vulnerability description"},{"category":"summary","text":"pillow: Pillow: Denial of Service via invalid pointer free during font rendering","title":"Vulnerability summary"},{"category":"other","text":"Red Hat rates this vulnerability as having an Important severity because an unauthenticated remote attacker could cause memory corruption or terminate the application process without requiring privileges, leading to a denial of service. Across Red Hat product environments, Pillow operates as an underlying imaging component in automated services and containerized workloads. The vulnerability is only triggered if an application accepts untrusted font files from users and renders text using non-zero stroke widths, leaving services that utilize standard system fonts unaffected.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-112893"},{"category":"external","summary":"RHBZ#2555878","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2555878"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-112893","url":"https://www.cve.org/CVERecord?id=CVE-2026-112893"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-112893","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-112893"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/commit/a7416de7f3a43128179d2cf68df3240ea8dafe13","url":"https://github.com/python-pillow/Pillow/commit/a7416de7f3a43128179d2cf68df3240ea8dafe13"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0","url":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-j4m9-6w36-pw47","url":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-j4m9-6w36-pw47"}],"release_date":"2026-10-19T21:53:49.979000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"},{"category":"workaround","details":"--- CME Mitigation Guidance (2026-10-20) ---\nDisclaimer: The following mitigation solutions are provided as a courtesy, pending the application of an official patch. This content was generated utilizing AI tools in conjunction with data from https://cmetaxonomy.org. Customers are advised to independently evaluate and test these temporary mitigations within their own infrastructure.\n\nThe highest-value action is to ensure glyph pointers are initialized to NULL and verified before invocation in FreeType stroked font rendering cleanup: upgrade to Pillow 13.0.0 (or the fixed Python Pillow/Python3.11 Pillow/Rhoai/Odh Openvino Model Server Rhel/Rhelai3/Bootc Cuda Rhel/Rhelai3/Bootc Gaudi Rhel/Rhelai3/Bootc Rocm Rhel/Rhelai3/Disk Image Cuda Rhel/Python3X Pillow/Rhaiis/Vllm Spyre Rhel/Rhoai/Odh Kserve Storage Initializer Rhel/Rhoai/Odh Kserve Agent Rhel/Rhoai/Odh Kserve Controller Rhel/Rhoai/Odh Kserve Router Rhel/Rhoai/Odh Feature Server Rhel/Rhoai/Odh Trustyai Nemo Guardrails Server Rhel/Rhaiis/Vllm Tpu Rhel/Rhoai/Odh Training Cuda128 Torch29 Py312 Rhel/Rhoai/Odh Training Rocm64 Torch29 Py312 Rhel/Rhaiis/Vllm Cpu Rhel/Rhaiis/Vllm Neuron Rhel/Ansible Automation Platform 25/Lightspeed Chatbot Rhel/Openshift Lightspeed/Lightspeed Ocp Rag Rhel/Rhoai/Odh Th06 Cpu Torch291 Py312 Rhel/Rhoai/Odh Th06 Cuda130 Torch291 Py312 Rhel/Rhoai/Odh Vllm Gaudi Rhel/Python3.12 Pillow/Rhaii/Vllm Gaudi Rhel/Rhaii/Vllm Neuron Rhel/Rhaii/Vllm Tpu Rhel/Lightspeed Core/Lightspeed Stack Rhel/Rhoai/Odh Kserve Autogluon Server Rhel/Rhai Early Access/Docling Sdk Cuda Rhel/Rhai Early Access/Docling Serve Cuda Rhel/Rhoai/Odh Llm D Kv Cache Rhel/Rhoai/Odh Th06 Cpu Torch210 Py312 Rhel/Rhoai/Odh Th06 Cuda130 Torch210 Py312 Rhel/Rhoai/Odh Th06 Rocm64 Torch291 Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Llmcompressor Cuda Py312 Rhel/Ask Sre/Rhel Cla/Rlsapi Rhel/Rhaiis/Model Opt Cuda Rhel/Rhaiis/Vllm Cuda Rhel/Rhaiis/Vllm Rocm Rhel/Rhaii/Model Opt Cuda Rhel/Rhaii/Vllm Cpu Rhel/Rhaii/Vllm Cuda Rhel/Rhaii/Vllm Rocm Rhel/Rhaii/Vllm Spyre Rhel/Ansible Automation Platform/Automation Dashboard Rhel/Ansible Automation Platform 26/Hub Rhel/Ansible Automation Platform 26/Lightspeed Chatbot Rhel/Ansible Automation Platform 27/Hub Rhel/Ansible Automation Platform 27/Lightspeed Chatbot Rhel/Galaxy Ng/Automation Hub Container/Aap Rag Content/Ansible Chatbot Stack/Koku Report Emailer/Exploit Intelligence/Vulnerability Analysis Rhel/Insights Host Inventory/Google Lightspeed Agent/Lscore Deploy/Lightspeed Core/Rag Tool Cpu Rhel/Lightspeed Core/Rag Tool Cuda 12.9 Rhel/Lightspeed Services Ocp/Ocp Rules Rhel/Openshift Lightspeed Tech Preview/Lightspeed Rag Tool Rhel/Openshift Lightspeed/Lightspeed Service Api Rhel/Rhoai/Odh Caikit Tgis Serving Rhel/Rhoai/Odh Llama Stack Core Rhel/Rhoai/Odh Pipeline Runtime Pytorch Llmcompressor Cuda Py312 Rhel/Rhoai/Odh Automl Rhel/Rhoai/Odh Autorag Rhel/Rhoai/Odh Mlflow Rhel/Rhoai/Odh Pipeline Runtime Datascience Cpu Py312 Rhel/Rhoai/Odh Pipeline Runtime Pytorch Cuda Py312 Rhel/Rhoai/Odh Pipeline Runtime Pytorch Rocm Py312 Rhel/Rhoai/Odh Pipeline Runtime Tensorflow Cuda Py312 Rhel/Rhoai/Odh Pipeline Runtime Tensorflow Rocm Py312 Rhel/Rhoai/Odh Trustyai Garak Lls Provider Dsp Rhel/Rhoai/Odh Workbench Codeserver Datascience Cpu Py312 Rhel/Rhoai/Odh Workbench Jupyter Datascience Cpu Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Cuda Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Rocm Py312 Rhel/Rhoai/Odh Workbench Jupyter Tensorflow Cuda Py312 Rhel/Rhoai/Odh Workbench Jupyter Tensorflow Rocm Py312 Rhel/Rhoai/Odh Workbench Jupyter Trustyai Cpu Py312 Rhel/Rhai/Guidellm Rhel/Rhoai/Odh Ogx Core Rhel/Rhoai/Odh Th Torch Cpu Py312 Rhel/Rhoai/Odh Th Torch Cuda Py312 Rhel/Rhoai/Odh Th Torch Rocm Py312 Rhel/Quay/Quay Service Tool/Quay/Quay Rhel/Rhelai3/Bootc Aws Cuda Rhel/Rhelai3/Bootc Azure Cuda Rhel/Rhelai3/Bootc Azure Rocm Rhel/Rhelai3/Bootc Gcp Cuda Rhel builds), which initialize the glyph pointer to NULL, reset it to NULL after freeing in the render loop, and check that glyph is not NULL before calling FT_Done_Glyph in cleanup, closing all 2 injection vectors (malformed font bytes supplied to FreeType rendering, crafted TrueType font triggering glyph loading failure). Until patched, this is a concern on applications accepting untrusted font data and rendering text using FreeType with a non-zero stroke width, so restrict applications from loading untrusted font files and avoid rendering FreeType text with a non-zero stroke width. These controls limit blast radius and contain exploitation, but they do not fix the underlying invalid pointer free defect — apply the upstream patch to fully remediate.","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"scores":[{"cvss_v3":{"attackComplexity":"LOW","attackVector":"NETWORK","availabilityImpact":"HIGH","baseScore":7.5,"baseSeverity":"HIGH","confidentialityImpact":"NONE","integrityImpact":"NONE","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"NONE","vectorString":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Important"}],"title":"pillow: Pillow: Denial of Service via invalid pointer free during font rendering"},{"cve":"CVE-2026-112894","cwe":{"id":"CWE-770","name":"Allocation of Resources Without Limits or Throttling"},"discovery_date":"2026-10-19T22:01:05.404464+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2555875"}],"notes":[{"category":"description","text":"A flaw was found in Pillow. When converting multi-frame images into a multi-page PDF document, the library repeatedly scans previously assigned object identifiers, resulting in quadratic processing time. A remote attacker could exploit this vulnerability by supplying a crafted image with a large frame count, causing excessive CPU consumption and a Denial of Service (DoS).","title":"Vulnerability description"},{"category":"summary","text":"pillow: pillow: Denial of Service via crafted multi-frame image conversion","title":"Vulnerability summary"},{"category":"other","text":"This vulnerability is rated as an Important severity flaw because an unauthenticated remote attacker can cause severe CPU exhaustion and service disruption by submitting minimal crafted inputs without needing system privileges. The elevated risk arises because even tiny files with declared high sequence numbers can tie up backend worker processes for prolonged periods in applications that convert user-provided graphics into multipage documents. Deployments remain unaffected if they only perform standard single-image rendering, validation, or do not expose multi-frame document export capabilities to untrusted users.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-112894"},{"category":"external","summary":"RHBZ#2555875","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2555875"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-112894","url":"https://www.cve.org/CVERecord?id=CVE-2026-112894"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-112894","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-112894"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/commit/f6a852555f1a1bdff7781abf25f1a3b8bd6ce6cb","url":"https://github.com/python-pillow/Pillow/commit/f6a852555f1a1bdff7781abf25f1a3b8bd6ce6cb"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0","url":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-jg86-v3jv-hrpm","url":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-jg86-v3jv-hrpm"}],"release_date":"2026-10-19T21:55:19.424000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"},{"category":"workaround","details":"--- CME Mitigation Guidance (2026-10-20) ---\nDisclaimer: The following mitigation solutions are provided as a courtesy, pending the application of an official patch. This content was generated utilizing AI tools in conjunction with data from https://cmetaxonomy.org. Customers are advised to independently evaluate and test these temporary mitigations within their own infrastructure.\n\nThe highest-value action is to validate and enforce upper bounds on image frame counts before passing untrusted files to Pillow multi-page PDF conversion with save_all enabled: upgrade to Pillow 13.0.0 (or the fixed Python Pillow/Python3.11 Pillow/Rhoai/Odh Openvino Model Server Rhel/Rhelai3/Bootc Cuda Rhel/Rhelai3/Bootc Gaudi Rhel/Rhelai3/Bootc Rocm Rhel/Rhelai3/Disk Image Cuda Rhel/Python3X Pillow/Rhaiis/Vllm Spyre Rhel/Rhoai/Odh Kserve Storage Initializer Rhel/Rhoai/Odh Kserve Agent Rhel/Rhoai/Odh Kserve Controller Rhel/Rhoai/Odh Kserve Router Rhel/Rhoai/Odh Feature Server Rhel/Rhoai/Odh Trustyai Nemo Guardrails Server Rhel/Rhaiis/Vllm Tpu Rhel/Rhoai/Odh Training Cuda128 Torch29 Py312 Rhel/Rhoai/Odh Training Rocm64 Torch29 Py312 Rhel/Rhaiis/Vllm Cpu Rhel/Rhaiis/Vllm Neuron Rhel/Ansible Automation Platform 25/Lightspeed Chatbot Rhel/Openshift Lightspeed/Lightspeed Ocp Rag Rhel/Rhoai/Odh Th06 Cpu Torch291 Py312 Rhel/Rhoai/Odh Th06 Cuda130 Torch291 Py312 Rhel/Rhoai/Odh Vllm Gaudi Rhel/Python3.12 Pillow/Rhaii/Vllm Gaudi Rhel/Rhaii/Vllm Neuron Rhel/Rhaii/Vllm Tpu Rhel/Lightspeed Core/Lightspeed Stack Rhel/Rhoai/Odh Kserve Autogluon Server Rhel/Rhai Early Access/Docling Sdk Cuda Rhel/Rhai Early Access/Docling Serve Cuda Rhel/Rhoai/Odh Llm D Kv Cache Rhel/Rhoai/Odh Th06 Cpu Torch210 Py312 Rhel/Rhoai/Odh Th06 Cuda130 Torch210 Py312 Rhel/Rhoai/Odh Th06 Rocm64 Torch291 Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Llmcompressor Cuda Py312 Rhel/Ask Sre/Rhel Cla/Rlsapi Rhel/Rhaiis/Model Opt Cuda Rhel/Rhaiis/Vllm Cuda Rhel/Rhaiis/Vllm Rocm Rhel/Rhaii/Model Opt Cuda Rhel/Rhaii/Vllm Cpu Rhel/Rhaii/Vllm Cuda Rhel/Rhaii/Vllm Rocm Rhel/Rhaii/Vllm Spyre Rhel/Ansible Automation Platform/Automation Dashboard Rhel/Ansible Automation Platform 26/Hub Rhel/Ansible Automation Platform 26/Lightspeed Chatbot Rhel/Ansible Automation Platform 27/Hub Rhel/Ansible Automation Platform 27/Lightspeed Chatbot Rhel/Galaxy Ng/Automation Hub Container/Aap Rag Content/Ansible Chatbot Stack/Koku Report Emailer/Exploit Intelligence/Vulnerability Analysis Rhel/Insights Host Inventory/Google Lightspeed Agent/Lscore Deploy/Lightspeed Core/Rag Tool Cpu Rhel/Lightspeed Core/Rag Tool Cuda 12.9 Rhel/Lightspeed Services Ocp/Ocp Rules Rhel/Openshift Lightspeed Tech Preview/Lightspeed Rag Tool Rhel/Openshift Lightspeed/Lightspeed Service Api Rhel/Rhoai/Odh Caikit Tgis Serving Rhel/Rhoai/Odh Llama Stack Core Rhel/Rhoai/Odh Pipeline Runtime Pytorch Llmcompressor Cuda Py312 Rhel/Rhoai/Odh Automl Rhel/Rhoai/Odh Autorag Rhel/Rhoai/Odh Mlflow Rhel/Rhoai/Odh Pipeline Runtime Datascience Cpu Py312 Rhel/Rhoai/Odh Pipeline Runtime Pytorch Cuda Py312 Rhel/Rhoai/Odh Pipeline Runtime Pytorch Rocm Py312 Rhel/Rhoai/Odh Pipeline Runtime Tensorflow Cuda Py312 Rhel/Rhoai/Odh Pipeline Runtime Tensorflow Rocm Py312 Rhel/Rhoai/Odh Trustyai Garak Lls Provider Dsp Rhel/Rhoai/Odh Workbench Codeserver Datascience Cpu Py312 Rhel/Rhoai/Odh Workbench Jupyter Datascience Cpu Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Cuda Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Rocm Py312 Rhel/Rhoai/Odh Workbench Jupyter Tensorflow Cuda Py312 Rhel/Rhoai/Odh Workbench Jupyter Tensorflow Rocm Py312 Rhel/Rhoai/Odh Workbench Jupyter Trustyai Cpu Py312 Rhel/Rhai/Guidellm Rhel/Rhoai/Odh Ogx Core Rhel/Rhoai/Odh Th Torch Cpu Py312 Rhel/Rhoai/Odh Th Torch Cuda Py312 Rhel/Rhoai/Odh Th Torch Rocm Py312 Rhel/Quay/Quay Service Tool/Quay/Quay Rhel/Rhelai3/Bootc Aws Cuda Rhel/Rhelai3/Bootc Azure Cuda Rhel/Rhelai3/Bootc Azure Rocm Rhel/Rhelai3/Bootc Gcp Cuda Rhel builds), which maintain a running maximum object ID attribute on XrefTable updated upon key insertion to resolve identifiers in constant time instead of scanning all keys with max(), closing all 2 injection vectors (crafted multi-frame image with a large number of frames, crafted small APNG image with an inflated declared frame count). Until patched, this is a concern on applications converting untrusted user-supplied multi-frame images to PDF documents using save_all=True across all platforms, so Restrict untrusted image uploads, enforce strict frame count validation before processing, or disable multi-page PDF conversion using save_all=True. As a compensating measure, Impose per-process CPU time limits using ulimit or prlimit (RLIMIT_CPU) to terminate runaway PDF conversion processes and bound CPU starvation, Configure service watchdog timers such as systemd WatchdogSec and Restart to restart processes that hang during quadratic object scanning, and Constrain cgroups v2 CPU quotas to restrict the blast radius of excessive CPU utilization during multi-frame PDF generation. These controls limit blast radius and contain exploitation, but they do not fix the underlying quadratic object identifier scanning defect — apply the upstream patch to fully remediate.\n\nHighest-impact CME: CME-1311 Input Size and Quantity Bounds Enforcement (high)","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"scores":[{"cvss_v3":{"attackComplexity":"LOW","attackVector":"NETWORK","availabilityImpact":"HIGH","baseScore":7.5,"baseSeverity":"HIGH","confidentialityImpact":"NONE","integrityImpact":"NONE","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"NONE","vectorString":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Important"}],"title":"pillow: pillow: Denial of Service via crafted multi-frame image conversion"},{"cve":"CVE-2026-112895","cwe":{"id":"CWE-835","name":"Loop with Unreachable Exit Condition"},"discovery_date":"2026-10-19T22:01:20.951322+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2555879"}],"notes":[{"category":"description","text":"A flaw was found in Pillow. This vulnerability allows a remote attacker to cause a Denial of Service (DoS) by supplying a specially crafted PDF document to an application performing PDF append operations. Because the PDF parser fails to properly track cyclic or deeply nested object references, processing the malicious document can consume excessive CPU and memory resources or trigger an application crash.","title":"Vulnerability description"},{"category":"summary","text":"pillow: pillow: Denial of service via crafted PDF during append operations","title":"Vulnerability summary"},{"category":"other","text":"This vulnerability is rated as an Important severity flaw because an unauthenticated remote attacker can cause severe service disruption and denial of service without requiring elevated privileges. Red Hat products shipping Pillow—including components within Red Hat OpenShift AI, Ansible Automation Platform, and Satellite—are exposed only when configured to append images to untrusted, user-supplied PDF documents. Standard deployments that do not append to external PDF files or that generate new documents from trusted sources are unaffected.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-112895"},{"category":"external","summary":"RHBZ#2555879","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2555879"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-112895","url":"https://www.cve.org/CVERecord?id=CVE-2026-112895"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-112895","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-112895"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/commit/e2ae7eacad6de68c37eb28b633a97bebacf2c5bf","url":"https://github.com/python-pillow/Pillow/commit/e2ae7eacad6de68c37eb28b633a97bebacf2c5bf"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0","url":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-p4wg-p73j-c528","url":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-p4wg-p73j-c528"}],"release_date":"2026-10-19T21:57:14.207000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"},{"category":"workaround","details":"--- CME Mitigation Guidance (2026-10-20) ---\nDisclaimer: The following mitigation solutions are provided as a courtesy, pending the application of an official patch. This content was generated utilizing AI tools in conjunction with data from https://cmetaxonomy.org. Customers are advised to independently evaluate and test these temporary mitigations within their own infrastructure.\n\nThe highest-value action is to disable PDF append operations or prevent untrusted PDF inputs from reaching Pillow Image.save calls: upgrade to Pillow 13.0.0 (or the fixed Python Pillow/Python3.11 Pillow/Rhoai/Odh Openvino Model Server Rhel/Rhelai3/Bootc Cuda Rhel/Rhelai3/Bootc Gaudi Rhel/Rhelai3/Bootc Rocm Rhel/Rhelai3/Disk Image Cuda Rhel/Python3X Pillow/Rhaiis/Vllm Spyre Rhel/Rhoai/Odh Kserve Storage Initializer Rhel/Rhoai/Odh Kserve Agent Rhel/Rhoai/Odh Kserve Controller Rhel/Rhoai/Odh Kserve Router Rhel/Rhoai/Odh Feature Server Rhel/Rhoai/Odh Trustyai Nemo Guardrails Server Rhel/Rhaiis/Vllm Tpu Rhel/Rhoai/Odh Training Cuda128 Torch29 Py312 Rhel/Rhoai/Odh Training Rocm64 Torch29 Py312 Rhel/Rhaiis/Vllm Cpu Rhel/Rhaiis/Vllm Neuron Rhel/Ansible Automation Platform 25/Lightspeed Chatbot Rhel/Openshift Lightspeed/Lightspeed Ocp Rag Rhel/Rhoai/Odh Th06 Cpu Torch291 Py312 Rhel/Rhoai/Odh Th06 Cuda130 Torch291 Py312 Rhel/Rhoai/Odh Vllm Gaudi Rhel/Python3.12 Pillow/Rhaii/Vllm Gaudi Rhel/Rhaii/Vllm Neuron Rhel/Rhaii/Vllm Tpu Rhel/Lightspeed Core/Lightspeed Stack Rhel/Rhoai/Odh Kserve Autogluon Server Rhel/Rhai Early Access/Docling Sdk Cuda Rhel/Rhai Early Access/Docling Serve Cuda Rhel/Rhoai/Odh Llm D Kv Cache Rhel/Rhoai/Odh Th06 Cpu Torch210 Py312 Rhel/Rhoai/Odh Th06 Cuda130 Torch210 Py312 Rhel/Rhoai/Odh Th06 Rocm64 Torch291 Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Llmcompressor Cuda Py312 Rhel/Ask Sre/Rhel Cla/Rlsapi Rhel/Rhaiis/Model Opt Cuda Rhel/Rhaiis/Vllm Cuda Rhel/Rhaiis/Vllm Rocm Rhel/Rhaii/Model Opt Cuda Rhel/Rhaii/Vllm Cpu Rhel/Rhaii/Vllm Cuda Rhel/Rhaii/Vllm Rocm Rhel/Rhaii/Vllm Spyre Rhel/Ansible Automation Platform/Automation Dashboard Rhel/Ansible Automation Platform 26/Hub Rhel/Ansible Automation Platform 26/Lightspeed Chatbot Rhel/Ansible Automation Platform 27/Hub Rhel/Ansible Automation Platform 27/Lightspeed Chatbot Rhel/Galaxy Ng/Automation Hub Container/Aap Rag Content/Ansible Chatbot Stack/Koku Report Emailer/Exploit Intelligence/Vulnerability Analysis Rhel/Insights Host Inventory/Google Lightspeed Agent/Lscore Deploy/Lightspeed Core/Rag Tool Cpu Rhel/Lightspeed Core/Rag Tool Cuda 12.9 Rhel/Lightspeed Services Ocp/Ocp Rules Rhel/Openshift Lightspeed Tech Preview/Lightspeed Rag Tool Rhel/Openshift Lightspeed/Lightspeed Service Api Rhel/Rhoai/Odh Caikit Tgis Serving Rhel/Rhoai/Odh Llama Stack Core Rhel/Rhoai/Odh Pipeline Runtime Pytorch Llmcompressor Cuda Py312 Rhel/Rhoai/Odh Automl Rhel/Rhoai/Odh Autorag Rhel/Rhoai/Odh Mlflow Rhel/Rhoai/Odh Pipeline Runtime Datascience Cpu Py312 Rhel/Rhoai/Odh Pipeline Runtime Pytorch Cuda Py312 Rhel/Rhoai/Odh Pipeline Runtime Pytorch Rocm Py312 Rhel/Rhoai/Odh Pipeline Runtime Tensorflow Cuda Py312 Rhel/Rhoai/Odh Pipeline Runtime Tensorflow Rocm Py312 Rhel/Rhoai/Odh Trustyai Garak Lls Provider Dsp Rhel/Rhoai/Odh Workbench Codeserver Datascience Cpu Py312 Rhel/Rhoai/Odh Workbench Jupyter Datascience Cpu Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Cuda Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Rocm Py312 Rhel/Rhoai/Odh Workbench Jupyter Tensorflow Cuda Py312 Rhel/Rhoai/Odh Workbench Jupyter Tensorflow Rocm Py312 Rhel/Rhoai/Odh Workbench Jupyter Trustyai Cpu Py312 Rhel/Rhai/Guidellm Rhel/Rhoai/Odh Ogx Core Rhel/Rhoai/Odh Th Torch Cpu Py312 Rhel/Rhoai/Odh Th Torch Cuda Py312 Rhel/Rhoai/Odh Th Torch Rocm Py312 Rhel/Quay/Quay Service Tool/Quay/Quay Rhel/Rhelai3/Bootc Aws Cuda Rhel/Rhelai3/Bootc Azure Cuda Rhel/Rhelai3/Bootc Azure Rocm Rhel/Rhelai3/Bootc Gcp Cuda Rhel builds), which replaces recursive page-tree traversal with an iterative stack of iterators and tracks visited object IDs in a set to reject duplicate or cyclic references, closing all 4 injection vectors (cyclic PDF page-tree child references, deeply nested PDF page trees, repeated non-cyclic /Kids references, long chains of unique PDF trailer offsets). Until patched, this is a concern on Applications accepting user-supplied PDF documents and invoking Image.save with PDF append enabled, so restrict user-supplied PDF uploads and audit application workflows to prevent untrusted files from being processed by Image.save with append=True. As a compensating measure, Enforce CPU and memory limits using ulimit or prlimit to cap resource consumption and contain host-level exhaustion, Deploy systemd service watchdogs with WatchdogSec and Restart directives to detect application hangs or process termination and restart services automatically, Enforce strict bounds on uploaded file size and structure complexity at the application boundary before passing data to Pillow's append parser, and Constrain process CPU and memory utilization using cgroups v2 to limit the blast radius of runaway loop traversal during PDF append operations. These controls limit blast radius and contain exploitation, but they do not fix the underlying uncontrolled recursion defect — apply the upstream patch to fully remediate.\n\nHighest-impact CME: CME-114 Process Resource Limits (ulimit/prlimit) (medium)","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"scores":[{"cvss_v3":{"attackComplexity":"LOW","attackVector":"NETWORK","availabilityImpact":"HIGH","baseScore":7.5,"baseSeverity":"HIGH","confidentialityImpact":"NONE","integrityImpact":"NONE","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"NONE","vectorString":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Important"}],"title":"pillow: pillow: Denial of service via crafted PDF during append operations"},{"cve":"CVE-2026-112896","cwe":{"id":"CWE-125","name":"Out-of-bounds Read"},"discovery_date":"2026-10-19T22:21:29.226245+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2555890"}],"notes":[{"category":"description","text":"A flaw was found in Pillow. When saving or re-encoding a Tagged Image File Format (TIFF) image containing inconsistent color map metadata, the encoder reads beyond an allocated memory buffer. A remote attacker could exploit this vulnerability by submitting a specially crafted image, leading to a Denial of Service (DoS) via an application crash or information disclosure of heap memory.","title":"Vulnerability description"},{"category":"summary","text":"pillow: Pillow: Denial of Service via out-of-bounds read in TIFF encoder","title":"Vulnerability summary"},{"category":"other","text":"This vulnerability is rated as Important severity for Red Hat products because remote, unauthenticated attackers can supply crafted image files over network endpoints to cause process termination or heap memory disclosure without requiring privileges. The impact primarily concerns applications and containerized services that ingest untrusted image data and subsequently re-encode or save it using the libtiff-backed encoder while retaining metadata. Environments and workflows that only read, validate, or display images without writing them back out to TIFF format are not affected.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-112896"},{"category":"external","summary":"RHBZ#2555890","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2555890"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-112896","url":"https://www.cve.org/CVERecord?id=CVE-2026-112896"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-112896","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-112896"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/commit/d03dc3c587a9b0bdc9e4304d70d59a635a87b742","url":"https://github.com/python-pillow/Pillow/commit/d03dc3c587a9b0bdc9e4304d70d59a635a87b742"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0","url":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-pchr-8j38-3x4p","url":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-pchr-8j38-3x4p"}],"release_date":"2026-10-19T21:59:17.743000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"},{"category":"workaround","details":"--- CME Mitigation Guidance (2026-10-20) ---\nDisclaimer: The following mitigation solutions are provided as a courtesy, pending the application of an official patch. This content was generated utilizing AI tools in conjunction with data from https://cmetaxonomy.org. Customers are advised to independently evaluate and test these temporary mitigations within their own infrastructure.\n\nThe highest-value action is to validate ColorMap tag length against BitsPerSample before passing retained metadata to the libtiff encoder during TIFF saves: upgrade to Pillow 13.0.0 (or the fixed Python Pillow/Python3.11 Pillow/Rhoai/Odh Openvino Model Server Rhel/Rhelai3/Bootc Cuda Rhel/Rhelai3/Bootc Gaudi Rhel/Rhelai3/Bootc Rocm Rhel/Rhelai3/Disk Image Cuda Rhel/Python3X Pillow/Rhaiis/Vllm Spyre Rhel/Rhoai/Odh Kserve Storage Initializer Rhel/Rhoai/Odh Kserve Agent Rhel/Rhoai/Odh Kserve Controller Rhel/Rhoai/Odh Kserve Router Rhel/Rhoai/Odh Feature Server Rhel/Rhoai/Odh Trustyai Nemo Guardrails Server Rhel/Rhaiis/Vllm Tpu Rhel/Rhoai/Odh Training Cuda128 Torch29 Py312 Rhel/Rhoai/Odh Training Rocm64 Torch29 Py312 Rhel/Rhaiis/Vllm Cpu Rhel/Rhaiis/Vllm Neuron Rhel/Ansible Automation Platform 25/Lightspeed Chatbot Rhel/Openshift Lightspeed/Lightspeed Ocp Rag Rhel/Rhoai/Odh Th06 Cpu Torch291 Py312 Rhel/Rhoai/Odh Th06 Cuda130 Torch291 Py312 Rhel/Rhoai/Odh Vllm Gaudi Rhel/Python3.12 Pillow/Rhaii/Vllm Gaudi Rhel/Rhaii/Vllm Neuron Rhel/Rhaii/Vllm Tpu Rhel/Lightspeed Core/Lightspeed Stack Rhel/Rhoai/Odh Kserve Autogluon Server Rhel/Rhai Early Access/Docling Sdk Cuda Rhel/Rhai Early Access/Docling Serve Cuda Rhel/Rhoai/Odh Llm D Kv Cache Rhel/Rhoai/Odh Th06 Cpu Torch210 Py312 Rhel/Rhoai/Odh Th06 Cuda130 Torch210 Py312 Rhel/Rhoai/Odh Th06 Rocm64 Torch291 Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Llmcompressor Cuda Py312 Rhel/Ask Sre/Rhel Cla/Rlsapi Rhel/Rhaiis/Model Opt Cuda Rhel/Rhaiis/Vllm Cuda Rhel/Rhaiis/Vllm Rocm Rhel/Rhaii/Model Opt Cuda Rhel/Rhaii/Vllm Cpu Rhel/Rhaii/Vllm Cuda Rhel/Rhaii/Vllm Rocm Rhel/Rhaii/Vllm Spyre Rhel/Ansible Automation Platform/Automation Dashboard Rhel/Ansible Automation Platform 26/Hub Rhel/Ansible Automation Platform 26/Lightspeed Chatbot Rhel/Ansible Automation Platform 27/Hub Rhel/Ansible Automation Platform 27/Lightspeed Chatbot Rhel/Galaxy Ng/Automation Hub Container/Aap Rag Content/Ansible Chatbot Stack/Koku Report Emailer/Exploit Intelligence/Vulnerability Analysis Rhel/Insights Host Inventory/Google Lightspeed Agent/Lscore Deploy/Lightspeed Core/Rag Tool Cpu Rhel/Lightspeed Core/Rag Tool Cuda 12.9 Rhel/Lightspeed Services Ocp/Ocp Rules Rhel/Openshift Lightspeed Tech Preview/Lightspeed Rag Tool Rhel/Openshift Lightspeed/Lightspeed Service Api Rhel/Rhoai/Odh Caikit Tgis Serving Rhel/Rhoai/Odh Llama Stack Core Rhel/Rhoai/Odh Pipeline Runtime Pytorch Llmcompressor Cuda Py312 Rhel/Rhoai/Odh Automl Rhel/Rhoai/Odh Autorag Rhel/Rhoai/Odh Mlflow Rhel/Rhoai/Odh Pipeline Runtime Datascience Cpu Py312 Rhel/Rhoai/Odh Pipeline Runtime Pytorch Cuda Py312 Rhel/Rhoai/Odh Pipeline Runtime Pytorch Rocm Py312 Rhel/Rhoai/Odh Pipeline Runtime Tensorflow Cuda Py312 Rhel/Rhoai/Odh Pipeline Runtime Tensorflow Rocm Py312 Rhel/Rhoai/Odh Trustyai Garak Lls Provider Dsp Rhel/Rhoai/Odh Workbench Codeserver Datascience Cpu Py312 Rhel/Rhoai/Odh Workbench Jupyter Datascience Cpu Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Cuda Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Rocm Py312 Rhel/Rhoai/Odh Workbench Jupyter Tensorflow Cuda Py312 Rhel/Rhoai/Odh Workbench Jupyter Tensorflow Rocm Py312 Rhel/Rhoai/Odh Workbench Jupyter Trustyai Cpu Py312 Rhel/Rhai/Guidellm Rhel/Rhoai/Odh Ogx Core Rhel/Rhoai/Odh Th Torch Cpu Py312 Rhel/Rhoai/Odh Th Torch Cuda Py312 Rhel/Rhoai/Odh Th Torch Rocm Py312 Rhel/Quay/Quay Service Tool/Quay/Quay Rhel/Rhelai3/Bootc Aws Cuda Rhel/Rhelai3/Bootc Azure Cuda Rhel/Rhelai3/Bootc Azure Rocm Rhel/Rhelai3/Bootc Gcp Cuda Rhel builds), which validate that the ColorMap tag length matches (1 << BitsPerSample) * 3 and compute plane stride dynamically from bits_per_sample instead of hardcoding a 256 stride, closing crafted TIFF image with inconsistent BitsPerSample and ColorMap length metadata. Until patched, this is a concern on applications processing untrusted TIFF files that re-encode or save them using the libtiff encoder, so avoid re-encoding untrusted TIFF files or strip retained ColorMap metadata prior to saving images with libtiff. These controls limit blast radius and contain exploitation, but they do not fix the underlying heap out-of-bounds read defect — apply the upstream patch to fully remediate.","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"scores":[{"cvss_v3":{"attackComplexity":"LOW","attackVector":"NETWORK","availabilityImpact":"HIGH","baseScore":8.2,"baseSeverity":"HIGH","confidentialityImpact":"LOW","integrityImpact":"NONE","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"NONE","vectorString":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:N/A:H","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Important"}],"title":"pillow: Pillow: Denial of Service via out-of-bounds read in TIFF encoder"},{"cve":"CVE-2026-112897","cwe":{"id":"CWE-908","name":"Use of Uninitialized Resource"},"discovery_date":"2026-10-19T22:21:12.210877+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2555888"}],"notes":[{"category":"description","text":"A flaw was found in Pillow. This flaw can lead to information disclosure when handling compressed Tagged Image File Format (TIFF) images. During incomplete decoding of image data, uninitialized buffer memory is not cleared before being converted into pixel rows. A remote attacker can exploit this vulnerability by submitting a crafted TIFF image to an application that decodes and returns or re-encodes the image, allowing them to view stale contents from heap memory.","title":"Vulnerability description"},{"category":"summary","text":"pillow: pillow: Information disclosure via uninitialized buffer in TIFF decoding","title":"Vulnerability summary"},{"category":"other","text":"Red Hat rates this vulnerability as Important because an unauthenticated remote attacker can exploit default image parsing workflows to extract sensitive process memory without requiring privileges or triggering application crashes. In Red Hat environments such as Red Hat OpenShift AI and Ansible Automation Platform, backend services that ingest user-provided images and return rendered thumbnails, previews, or re-encoded files inadvertently disclose previously freed heap data from concurrent or prior operations. The vulnerability requires processing compressed TIFF files where the decoder handles truncated image streams, meaning services that reject this format or never expose generated image outputs are not impacted.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-112897"},{"category":"external","summary":"RHBZ#2555888","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2555888"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-112897","url":"https://www.cve.org/CVERecord?id=CVE-2026-112897"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-112897","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-112897"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/commit/1d4773e6cea0c4a2814e6d4cdc099bbcf56da6ce","url":"https://github.com/python-pillow/Pillow/commit/1d4773e6cea0c4a2814e6d4cdc099bbcf56da6ce"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0","url":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-q45j-58j2-45rg","url":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-q45j-58j2-45rg"}],"release_date":"2026-10-19T22:00:52.749000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"},{"category":"workaround","details":"--- CME Mitigation Guidance (2026-10-20) ---\nDisclaimer: The following mitigation solutions are provided as a courtesy, pending the application of an official patch. This content was generated utilizing AI tools in conjunction with data from https://cmetaxonomy.org. Customers are advised to independently evaluate and test these temporary mitigations within their own infrastructure.\n\nThe highest-value action is to zero uninitialized heap buffer memory during Pillow TIFF strip and tile decoding before exposing pixel rows to callers: upgrade to Pillow 13.0.0 (or the fixed Python Pillow/Python3.11 Pillow/Rhoai/Odh Openvino Model Server Rhel/Rhelai3/Bootc Cuda Rhel/Rhelai3/Bootc Gaudi Rhel/Rhelai3/Bootc Rocm Rhel/Rhelai3/Disk Image Cuda Rhel/Python3X Pillow/Rhaiis/Vllm Spyre Rhel/Rhoai/Odh Kserve Storage Initializer Rhel/Rhoai/Odh Kserve Agent Rhel/Rhoai/Odh Kserve Controller Rhel/Rhoai/Odh Kserve Router Rhel/Rhoai/Odh Feature Server Rhel/Rhoai/Odh Trustyai Nemo Guardrails Server Rhel/Rhaiis/Vllm Tpu Rhel/Rhoai/Odh Training Cuda128 Torch29 Py312 Rhel/Rhoai/Odh Training Rocm64 Torch29 Py312 Rhel/Rhaiis/Vllm Cpu Rhel/Rhaiis/Vllm Neuron Rhel/Ansible Automation Platform 25/Lightspeed Chatbot Rhel/Openshift Lightspeed/Lightspeed Ocp Rag Rhel/Rhoai/Odh Th06 Cpu Torch291 Py312 Rhel/Rhoai/Odh Th06 Cuda130 Torch291 Py312 Rhel/Rhoai/Odh Vllm Gaudi Rhel/Python3.12 Pillow/Rhaii/Vllm Gaudi Rhel/Rhaii/Vllm Neuron Rhel/Rhaii/Vllm Tpu Rhel/Lightspeed Core/Lightspeed Stack Rhel/Rhoai/Odh Kserve Autogluon Server Rhel/Rhai Early Access/Docling Sdk Cuda Rhel/Rhai Early Access/Docling Serve Cuda Rhel/Rhoai/Odh Llm D Kv Cache Rhel/Rhoai/Odh Th06 Cpu Torch210 Py312 Rhel/Rhoai/Odh Th06 Cuda130 Torch210 Py312 Rhel/Rhoai/Odh Th06 Rocm64 Torch291 Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Llmcompressor Cuda Py312 Rhel/Ask Sre/Rhel Cla/Rlsapi Rhel/Rhaiis/Model Opt Cuda Rhel/Rhaiis/Vllm Cuda Rhel/Rhaiis/Vllm Rocm Rhel/Rhaii/Model Opt Cuda Rhel/Rhaii/Vllm Cpu Rhel/Rhaii/Vllm Cuda Rhel/Rhaii/Vllm Rocm Rhel/Rhaii/Vllm Spyre Rhel/Ansible Automation Platform/Automation Dashboard Rhel/Ansible Automation Platform 26/Hub Rhel/Ansible Automation Platform 26/Lightspeed Chatbot Rhel/Ansible Automation Platform 27/Hub Rhel/Ansible Automation Platform 27/Lightspeed Chatbot Rhel/Galaxy Ng/Automation Hub Container/Aap Rag Content/Ansible Chatbot Stack/Koku Report Emailer/Exploit Intelligence/Vulnerability Analysis Rhel/Insights Host Inventory/Google Lightspeed Agent/Lscore Deploy/Lightspeed Core/Rag Tool Cpu Rhel/Lightspeed Core/Rag Tool Cuda 12.9 Rhel/Lightspeed Services Ocp/Ocp Rules Rhel/Openshift Lightspeed Tech Preview/Lightspeed Rag Tool Rhel/Openshift Lightspeed/Lightspeed Service Api Rhel/Rhoai/Odh Caikit Tgis Serving Rhel/Rhoai/Odh Llama Stack Core Rhel/Rhoai/Odh Pipeline Runtime Pytorch Llmcompressor Cuda Py312 Rhel/Rhoai/Odh Automl Rhel/Rhoai/Odh Autorag Rhel/Rhoai/Odh Mlflow Rhel/Rhoai/Odh Pipeline Runtime Datascience Cpu Py312 Rhel/Rhoai/Odh Pipeline Runtime Pytorch Cuda Py312 Rhel/Rhoai/Odh Pipeline Runtime Pytorch Rocm Py312 Rhel/Rhoai/Odh Pipeline Runtime Tensorflow Cuda Py312 Rhel/Rhoai/Odh Pipeline Runtime Tensorflow Rocm Py312 Rhel/Rhoai/Odh Trustyai Garak Lls Provider Dsp Rhel/Rhoai/Odh Workbench Codeserver Datascience Cpu Py312 Rhel/Rhoai/Odh Workbench Jupyter Datascience Cpu Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Cuda Py312 Rhel/Rhoai/Odh Workbench Jupyter Pytorch Rocm Py312 Rhel/Rhoai/Odh Workbench Jupyter Tensorflow Cuda Py312 Rhel/Rhoai/Odh Workbench Jupyter Tensorflow Rocm Py312 Rhel/Rhoai/Odh Workbench Jupyter Trustyai Cpu Py312 Rhel/Rhai/Guidellm Rhel/Rhoai/Odh Ogx Core Rhel/Rhoai/Odh Th Torch Cpu Py312 Rhel/Rhoai/Odh Th Torch Cuda Py312 Rhel/Rhoai/Odh Th Torch Rocm Py312 Rhel/Quay/Quay Service Tool/Quay/Quay Rhel/Rhelai3/Bootc Aws Cuda Rhel/Rhelai3/Bootc Azure Cuda Rhel/Rhelai3/Bootc Azure Rocm Rhel/Rhelai3/Bootc Gcp Cuda Rhel builds), which zero out newly expanded buffer memory beyond the decoded size with memset in _decodeTile and _decodeStrip, closing all 2 injection vectors (crafted JPEG-compressed TIFF data, crafted Group 4 fax compressed TIFF data). Until patched, this is a concern on Applications decoding user-supplied TIFF images with Pillow and returning, previewing, or re-encoding pixel data, so Restrict or block untrusted TIFF image submissions to applications that decode and return, preview, or re-encode pixel data. These controls limit blast radius and contain exploitation, but they do not fix the underlying uninitialized heap buffer defect — apply the upstream patch to fully remediate.","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"scores":[{"cvss_v3":{"attackComplexity":"LOW","attackVector":"NETWORK","availabilityImpact":"NONE","baseScore":7.5,"baseSeverity":"HIGH","confidentialityImpact":"HIGH","integrityImpact":"NONE","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"NONE","vectorString":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:N/A:N","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Important"}],"title":"pillow: pillow: Information disclosure via uninitialized buffer in TIFF decoding"},{"cve":"CVE-2026-112898","cwe":{"id":"CWE-908","name":"Use of Uninitialized Resource"},"discovery_date":"2026-10-19T22:21:02.306792+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2555887"}],"notes":[{"category":"description","text":"A flaw was found in Pillow. When processing an image with an empty palette during color quantization, internal stack memory is left uninitialized and emitted into the output image data. An attacker who supplies a crafted palette image can exploit this vulnerability to read uninitialized memory contents from the resulting image file, leading to information disclosure.","title":"Vulnerability description"},{"category":"summary","text":"pillow: Pillow: Information disclosure via empty palette quantization","title":"Vulnerability summary"},{"category":"other","text":"Red Hat rates this vulnerability as Moderate because it is confined to localized memory leakage and cannot be leveraged directly for service interruption or arbitrary code execution. Exploitation requires an application to ingest user-supplied color maps and pass them into quantization routines, then return the generated graphic to the requester. In standard Red Hat product deployments, integrated imaging tasks rarely allow external users to control color quantization palettes, thereby substantially restricting real-world attack vectors.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-112898"},{"category":"external","summary":"RHBZ#2555887","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2555887"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-112898","url":"https://www.cve.org/CVERecord?id=CVE-2026-112898"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-112898","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-112898"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/commit/13943c8f857b22f073d30f0ef9b725a0d255f1a4","url":"https://github.com/python-pillow/Pillow/commit/13943c8f857b22f073d30f0ef9b725a0d255f1a4"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0","url":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-rfgq-q3w2-pmv2","url":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-rfgq-q3w2-pmv2"}],"release_date":"2026-10-19T22:04:23.462000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"}],"scores":[{"cvss_v3":{"attackComplexity":"LOW","attackVector":"NETWORK","availabilityImpact":"NONE","baseScore":5.3,"baseSeverity":"MEDIUM","confidentialityImpact":"LOW","integrityImpact":"NONE","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"NONE","vectorString":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:N/A:N","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Moderate"}],"title":"pillow: Pillow: Information disclosure via empty palette quantization"},{"cve":"CVE-2026-112899","cwe":{"id":"CWE-409","name":"Improper Handling of Highly Compressed Data (Data Amplification)"},"discovery_date":"2026-11-03T00:00:00+00:00","notes":[{"category":"description","text":"A flaw was found in Pillow. The Encapsulated PostScript (EPS) image loader only validates decompression bomb limits against the unscaled dimensions defined in the file, rather than the final scaled dimensions requested during rendering. An attacker can exploit this issue by providing a crafted EPS file, forcing the rendering process to generate excessively large temporary files and consume excessive CPU and disk resources, resulting in a Denial of Service (DoS).","title":"Vulnerability description"},{"category":"summary","text":"Pillow: EPS load(scale) lets Ghostscript render past the decompression bomb limit","title":"Vulnerability summary"},{"category":"other","text":"This vulnerability is rated as Moderate severity because it results in a denial of service through excessive CPU and storage consumption without compromising system confidentiality or integrity. In Red Hat environments, exploitation requires an application using the Pillow library to ingest untrusted Encapsulated PostScript (EPS) files and perform scaled rendering via Ghostscript. Because the risk is confined to availability exhaustion and does not allow remote code execution or privilege escalation, the blast radius is limited.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-112899"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-112899","url":"https://www.cve.org/CVERecord?id=CVE-2026-112899"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-112899","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-112899"}],"release_date":"2026-10-20T00:00:00+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"},{"category":"workaround","details":"Operational controls can mitigate exposure to this vulnerability:\n\n1. Restrict file formats: Configure applications using Pillow to validate and accept only required raster formats (such as PNG or JPEG), preventing untrusted EPS files from being processed.\n2. Remove the Ghostscript rendering backend: If PostScript or EPS rendering is not required, remove the Ghostscript package to prevent Pillow from delegating rendering to it:\n\n# dnf remove ghostscript\n\n3. Sandbox and resource limits: Run image processing workloads in containerized or sandboxed environments with strict disk quota and CPU limits to prevent excessive temporary file generation from exhausting host resources.\n\nCaveats: Removing Ghostscript will disable EPS and PostScript rendering capabilities across the system for any applications that depend on Ghostscript.","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"scores":[{"cvss_v3":{"attackComplexity":"LOW","attackVector":"NETWORK","availabilityImpact":"HIGH","baseScore":7.5,"baseSeverity":"HIGH","confidentialityImpact":"NONE","integrityImpact":"NONE","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"NONE","vectorString":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Moderate"}],"title":"Pillow: EPS load(scale) lets Ghostscript render past the decompression bomb limit"},{"cve":"CVE-2026-112900","cwe":{"id":"CWE-88","name":"Improper Neutralization of Argument Delimiters in a Command"},"discovery_date":"2026-10-19T22:21:21.457146+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2555889"}],"notes":[{"category":"description","text":"A flaw was found in Pillow. This vulnerability may allow an attacker to achieve arbitrary code execution when processing Encapsulated PostScript (EPS) images. Because Pillow invokes the external Ghostscript utility without properly sanitizing user-controlled input filenames, names beginning with specific command prefixes are misinterpreted as options rather than file paths. An attacker who controls the filename and can supply an additional file containing command arguments can exploit this argument injection to execute arbitrary commands on the system.","title":"Vulnerability description"},{"category":"summary","text":"pillow: Pillow: Argument Injection in Ghostscript call, potentially leading into RCE","title":"Vulnerability summary"},{"category":"other","text":"This vulnerability is rated Critical because an attacker capable of supplying an unvalidated filename to image handling services can achieve arbitrary remote code execution via argument injection into Ghostscript. In Red Hat environments, exploitation requires the host application to process Encapsulated PostScript files, have the optional Ghostscript package installed, and allow an attacker to place an auxiliary argument file on the target system. Workloads that do not accept untrusted filenames or that run in environments without the external Ghostscript binary are not susceptible to this attack vector.","title":"Statement"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-112900"},{"category":"external","summary":"RHBZ#2555889","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2555889"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-112900","url":"https://www.cve.org/CVERecord?id=CVE-2026-112900"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-112900","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-112900"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/commit/a2c363141da31a265dd536c945677db4b64f3ec9","url":"https://github.com/python-pillow/Pillow/commit/a2c363141da31a265dd536c945677db4b64f3ec9"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0","url":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-w889-685w-c8fh","url":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-w889-685w-c8fh"}],"release_date":"2026-10-19T22:07:50.418000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"}],"threats":[{"category":"impact","details":"Critical"}],"title":"pillow: Pillow: Argument Injection in Ghostscript call, potentially leading into RCE"},{"cve":"CVE-2026-112901","cwe":{"id":"CWE-1333","name":"Inefficient Regular Expression Complexity"},"discovery_date":"2026-10-19T22:21:54.134943+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2555891"}],"notes":[{"category":"description","text":"A flaw was found in Pillow. A remote attacker can cause a Denial of Service (DoS) by supplying a specially crafted XBM (X BitMap) image file. When the application parses the image header, an inefficient regular expression causes excessive computational backtracking, stalling worker processes and leaving the service unresponsive.","title":"Vulnerability description"},{"category":"summary","text":"pillow: Pillow: Denial of Service via crafted XBM image header","title":"Vulnerability summary"},{"category":"other","text":"This vulnerability is rated as Moderate impact because an attacker can cause CPU exhaustion and stall individual service worker processes by submitting a malformed image, without compromising system confidentiality or integrity. In Red Hat environments, Pillow is embedded across multiple platform services and containerized tools to handle graphic operations. Real-world impact is largely confined to endpoints that ingest untrusted image uploads without enforcing strict format restrictions.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-112901"},{"category":"external","summary":"RHBZ#2555891","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2555891"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-112901","url":"https://www.cve.org/CVERecord?id=CVE-2026-112901"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-112901","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-112901"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/commit/2852a3e3d568997ddec745fc800991f138878f4b","url":"https://github.com/python-pillow/Pillow/commit/2852a3e3d568997ddec745fc800991f138878f4b"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0","url":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-xgh8-jp6q-p5x8","url":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-xgh8-jp6q-p5x8"}],"release_date":"2026-10-19T22:09:52.614000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"}],"scores":[{"cvss_v3":{"attackComplexity":"LOW","attackVector":"NETWORK","availabilityImpact":"LOW","baseScore":5.3,"baseSeverity":"MEDIUM","confidentialityImpact":"NONE","integrityImpact":"NONE","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"NONE","vectorString":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:L","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Moderate"}],"title":"pillow: Pillow: Denial of Service via crafted XBM image header"},{"cve":"CVE-2026-112902","cwe":{"id":"CWE-125","name":"Out-of-bounds Read"},"discovery_date":"2026-11-03T05:24:58.017000+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2565602"}],"notes":[{"category":"description","text":"A flaw was found in Pillow. When saving an image in the TIFF format, Pillow incorrectly forwards unknown metadata tags to the underlying image processing library. A remote attacker could exploit this flaw by providing a specially crafted image containing malicious metadata to an application that processes and re-saves it. Successful exploitation allows the attacker to read out-of-bounds memory, leading to the disclosure of sensitive process memory or causing the application to crash, resulting in a Denial of Service (DoS).","title":"Vulnerability description"},{"category":"summary","text":"Pillow: Memory disclosure and segfault in libtiff encoder via crafted TIFF tags","title":"Vulnerability summary"},{"category":"other","text":"This vulnerability is rated as having Important severity because it enables unauthenticated remote attackers to obtain sensitive process memory or terminate worker processes by submitting specially crafted image files. In Red Hat environments, Pillow is commonly deployed in web services and automated pipelines to process and transform user-supplied graphics. Exposure is primarily limited to application workflows that ingest untrusted images and write them out using the TIFF format, where mismatched tag handling leads to out-of-bounds reads. Successful exploitation can defeat address space layout randomization (ASLR) or disrupt image ingestion services through persistent segmentation faults.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-112902"},{"category":"external","summary":"RHBZ#2565602","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2565602"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-112902","url":"https://www.cve.org/CVERecord?id=CVE-2026-112902"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-112902","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-112902"}],"release_date":"2026-10-20T00:00:00+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"},{"category":"workaround","details":"To mitigate this vulnerability, apply operational controls within applications utilizing Pillow:\n\n1. Restrict output formats: Configure image-processing workflows to avoid re-encoding or exporting untrusted user-submitted files into the TIFF format.\n2. Strip metadata: If converting images to TIFF is required, strip or omit untrusted metadata, such as EXIF data or custom tags, before invoking the image save routine.\n3. Process isolation: Isolate image rendering and conversion workloads within sandboxed environments or restricted containers with strict memory boundaries to limit access to sensitive memory and minimize the blast radius of worker crashes.\n\nCaveats: Restricting image export formats or stripping metadata may alter downstream image handling and remove required tag information from generated files.","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"scores":[{"cvss_v3":{"attackComplexity":"LOW","attackVector":"LOCAL","availabilityImpact":"HIGH","baseScore":7.1,"baseSeverity":"HIGH","confidentialityImpact":"HIGH","integrityImpact":"NONE","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"REQUIRED","vectorString":"CVSS:3.1/AV:L/AC:L/PR:N/UI:R/S:U/C:H/I:N/A:H","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Important"}],"title":"Pillow: Memory disclosure and segfault in libtiff encoder via crafted TIFF tags"},{"cve":"CVE-2026-112940","cwe":{"id":"CWE-367","name":"Time-of-check Time-of-use (TOCTOU) Race Condition"},"discovery_date":"2026-10-20T14:51:27.340888+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2556219"}],"notes":[{"category":"description","text":"A flaw was found in filelock. When utilizing the ReadWriteLock mechanism, the component reopens SQLite database and auxiliary files by file path rather than retaining secure file descriptors. A local attacker with file renaming permissions in a shared directory can swap these filenames during operation, causing the application to delete or overwrite unintended files. This issue can result in unauthorized data modification, data loss, or a Denial of Service (DoS).","title":"Vulnerability description"},{"category":"summary","text":"filelock: filelock: Arbitrary file overwrite via insecure database file reopening","title":"Vulnerability summary"},{"category":"other","text":"This vulnerability is rated as Moderate severity rather than Important because successful exploitation requires local account access and non-standard directory permissions where untrusted users can manipulate filenames within an unsticky parent directory. Under standard Red Hat Enterprise Linux installations, default kernel settings such as fs.protected_regular and fs.protected_hardlinks restrict link manipulation in common shared locations like /tmp. In addition, exposure is limited to workloads configured to utilize the SQLite-backed ReadWriteLock mechanism rather than standard single-process locking.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-112940"},{"category":"external","summary":"RHBZ#2556219","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2556219"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-112940","url":"https://www.cve.org/CVERecord?id=CVE-2026-112940"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-112940","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-112940"},{"category":"external","summary":"https://github.com/tox-dev/filelock/commit/5e3f2ec0c5fb17cb2635b60b2b68bc63ab81ca2d","url":"https://github.com/tox-dev/filelock/commit/5e3f2ec0c5fb17cb2635b60b2b68bc63ab81ca2d"},{"category":"external","summary":"https://github.com/tox-dev/filelock/pull/834","url":"https://github.com/tox-dev/filelock/pull/834"},{"category":"external","summary":"https://github.com/tox-dev/filelock/releases/tag/4.2.1","url":"https://github.com/tox-dev/filelock/releases/tag/4.2.1"},{"category":"external","summary":"https://github.com/tox-dev/filelock/security/advisories/GHSA-v4x8-3wr5-3wq8","url":"https://github.com/tox-dev/filelock/security/advisories/GHSA-v4x8-3wr5-3wq8"}],"release_date":"2026-10-20T14:41:49.555000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"}],"scores":[{"cvss_v3":{"attackComplexity":"HIGH","attackVector":"LOCAL","availabilityImpact":"HIGH","baseScore":6.3,"baseSeverity":"MEDIUM","confidentialityImpact":"NONE","integrityImpact":"HIGH","privilegesRequired":"LOW","scope":"UNCHANGED","userInteraction":"NONE","vectorString":"CVSS:3.1/AV:L/AC:H/PR:L/UI:N/S:U/C:N/I:H/A:H","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Moderate"}],"title":"filelock: filelock: Arbitrary file overwrite via insecure database file reopening"},{"cve":"CVE-2026-114406","cwe":{"id":"CWE-295","name":"Improper Certificate Validation"},"discovery_date":"2026-10-20T09:01:55.916000+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2556046"}],"notes":[{"category":"description","text":"The system certificate authority (CA) store in Red Hat products, distributed via ca-certificates RPM packages, included multiple root CA certificates of authorities, which signed intermediate CA certificates with weak private keys issued by the eMudhra certificate authority. A man-in-the-middle attacker could use this flaw to issue certificates for arbitrary host names signed by the weak intermediate CA certificates, which would be considered trusted by a TLS/SSL client using system CA store and cause such clients to connect to malicious sites.\n\nDue to inability to only distrust specific intermediate certificates in all clients using the system CA store, root CA certificates of authorities that signed weak intermediate CA certificates are being removed.","title":"Vulnerability description"},{"category":"summary","text":"ca-certificates: Distrust eMudhra intermediate CA certificates due to security concerns","title":"Vulnerability summary"},{"category":"other","text":"This CVE has been assigned to track trust store updates to distrust compromised certificates, but is not a vulnerability. For details, see RHSB-2026-012.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-114406"},{"category":"external","summary":"RHBZ#2556046","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2556046"},{"category":"external","summary":"RHSB-2026-012","url":"https://access.redhat.com/security/vulnerabilities/RHSB-2026-012"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-114406","url":"https://www.cve.org/CVERecord?id=CVE-2026-114406"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-114406","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-114406"}],"release_date":"2026-10-20T13:00:00+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"},{"category":"workaround","details":"See the security bulletin for a detailed mitigation procedure.","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"scores":[{"cvss_v3":{"attackComplexity":"HIGH","attackVector":"NETWORK","availabilityImpact":"NONE","baseScore":6.8,"baseSeverity":"MEDIUM","confidentialityImpact":"HIGH","integrityImpact":"HIGH","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"REQUIRED","vectorString":"CVSS:3.1/AV:N/AC:H/PR:N/UI:R/S:U/C:H/I:H/A:N","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Critical"}],"title":"ca-certificates: Distrust eMudhra intermediate CA certificates due to security concerns"},{"cve":"CVE-2026-114501","cwe":{"id":"CWE-908","name":"Use of Uninitialized Resource"},"discovery_date":"2026-10-21T21:41:26.337655+00:00","ids":[{"system_name":"Red Hat Bugzilla ID","text":"2558580"}],"notes":[{"category":"description","text":"A flaw was found in Pillow. This vulnerability can lead to information disclosure when exporting images to Apache Arrow format with memory alignment enabled, as the library assumes image rows are contiguous in memory. Consequently, exported buffers may incorporate uninitialized padding or residual data from previously freed process memory. An attacker could exploit this flaw to obtain sensitive process memory or image data belonging to other users.","title":"Vulnerability description"},{"category":"summary","text":"pillow: Pillow: Information disclosure via improper memory alignment during Arrow export","title":"Vulnerability summary"},{"category":"other","text":"This vulnerability is rated as having a Low impact because exploitation depends on non-default memory alignment configurations combined with image conversion to Apache Arrow, limiting exposure to specialized multi-tenant pipelines. Standard Red Hat software components and default workflows do not enable custom memory alignment during image transfers. Furthermore, the risk is strictly restricted to information disclosure of residual heap data or uninitialized padding within the process without providing arbitrary code execution or data modification capabilities.","title":"Statement"},{"category":"general","text":"The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.","title":"CVSS score applicability"}],"product_status":{"fixed":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]},"references":[{"category":"self","summary":"Canonical URL","url":"https://access.redhat.com/security/cve/CVE-2026-114501"},{"category":"external","summary":"RHBZ#2558580","url":"https://bugzilla.redhat.com/show_bug.cgi?id=2558580"},{"category":"external","summary":"https://www.cve.org/CVERecord?id=CVE-2026-114501","url":"https://www.cve.org/CVERecord?id=CVE-2026-114501"},{"category":"external","summary":"https://nvd.nist.gov/vuln/detail/CVE-2026-114501","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-114501"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/commit/19f64946eb247b973dd4560be7dea3bf979d2e6a","url":"https://github.com/python-pillow/Pillow/commit/19f64946eb247b973dd4560be7dea3bf979d2e6a"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/pull/10172","url":"https://github.com/python-pillow/Pillow/pull/10172"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0","url":"https://github.com/python-pillow/Pillow/releases/tag/13.0.0"},{"category":"external","summary":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-654x-cwwv-5j9c","url":"https://github.com/python-pillow/Pillow/security/advisories/GHSA-654x-cwwv-5j9c"}],"release_date":"2026-10-21T21:17:09.748000+00:00","remediations":[{"category":"vendor_fix","date":"2026-11-03T19:52:15+00:00","details":"For more information visit https://access.redhat.com/errata/RHSA-2026:94715","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"],"restart_required":{"category":"none"},"url":"https://access.redhat.com/errata/RHSA-2026:94715"},{"category":"workaround","details":"Mitigation for this issue is either not available or the currently available options do not meet the Red Hat Product Security criteria comprising ease of use and deployment, applicability to widespread installation base, or stability.","product_ids":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"scores":[{"cvss_v3":{"attackComplexity":"HIGH","attackVector":"NETWORK","availabilityImpact":"NONE","baseScore":3.7,"baseSeverity":"LOW","confidentialityImpact":"LOW","integrityImpact":"NONE","privilegesRequired":"NONE","scope":"UNCHANGED","userInteraction":"NONE","vectorString":"CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:L/I:N/A:N","version":"3.1"},"products":["Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:2209628d7049c8d1b45a043b3ac31d21cc879be1c6618b4849228399d6541776_amd64","Red Hat AI Inference Server 3.5:registry.redhat.io/rhaii/model-opt-cuda-rhel9@sha256:b81f3228c1ce2dfda30e4c83c111d4b671706e376566ab2d98f9326cb3852d16_arm64"]}],"threats":[{"category":"impact","details":"Low"}],"title":"pillow: Pillow: Information disclosure via improper memory alignment during Arrow export"}]}