{"resourceId":"rhsa-2026-86891-openshift-ai-352-critical-images-20261021","versions":[{"version":"external-2b5695e6d7c6d3ec839d6cac5bd1318f71861d3cfc0cc0cb97f6aebd3d038535","resource":{"id":"rhsa-2026-86891-openshift-ai-352-critical-images-20261021","title":"RHSA-2026:86891: Critical security advisory with updated images for Red Hat OpenShift AI 3.5.2","organization":"Red Hat Product Security","sector":"Strategic partner platform: OpenShift AI security maintenance","geography":"Global (self-managed OpenShift AI customers)","publishedAt":"Red Hat Customer Portal errata, issued Oct. 21, 2026","publicationDate":"2026-10-21","eventDate":"2026-10-21","sourceName":"Red Hat Customer Portal","sourceLabel":"RHSA-2026:86891 - Security Advisory","sourceUrl":"https://access.redhat.com/errata/RHSA-2026:86891","evidenceClass":"vendor-claim","outcomeClass":"cautionary","topics":["data-security","infrastructure","operating-model"],"finding":"Red Hat issued RHSA-2026:86891 on Oct. 21, 2026, rated Critical, providing updated container images for Red Hat OpenShift AI 3.5.2. The advisory lists many CVEs (including CVE-2024-14029 and numerous 2026 CVEs) and updated images for components including the AI gateway, AutoRAG, AutoML, EvalHub, dashboard, data science pipelines and Feast operators. Customers are directed to the OpenShift AI upgrade documentation to fully apply the errata.","sledRelevance":"State, local and university teams running self-managed OpenShift AI own patching. A Critical rating on the current 3.5.x line, spanning the components marketed for governed RAG and agents, means AI platform operations must follow the same vulnerability-management SLAs as other critical infrastructure.","evidence":"Opened the errata page and read the summary, severity, issued and updated dates (both 2026-10-21), the affected-product description, the CVE list and the image list. Individual CVE details were not opened, so which CVEs drive the Critical rating, and whether they are exploitable in typical configurations, is not established here.","architectureImplications":"OpenShift AI is delivered as many operator-managed images, so applying the fix requires an operator-driven upgrade, not a single package update. Disconnected (air-gapped) installs need image mirroring before upgrade.","governanceImplications":"AI platforms should be in the agency's vulnerability-management scope with defined patch windows. Upgrade and rollback testing belongs in change control.","securityPrivacyImplications":"The breadth of updated images (gateway, RAG, evaluation, pipelines) means unpatched clusters expose the components that handle prompts, documents and model access.","caveats":"A vendor advisory states severity and fixed images but not exploitability in specific deployments. CVE-level details were not reviewed for this edition.","streamIds":["red-hat"],"roles":{"sales":"Sales framing: use this to frame total cost of ownership honestly. Self-managed OpenShift AI gives agencies control over data and models, but they also own patching. A Critical advisory on the current release is a natural opening to discuss managed services, Technical Account Manager coverage or partner-run operations for SLED customers without deep Kubernetes staff. Do not imply the platform is less secure than alternatives. The point is that AI platforms need operational coverage.","engineering":"Engineering framing: confirm the customer's OpenShift AI version and update channel, then plan the operator upgrade to 3.5.2, with image mirroring for disconnected clusters. Because the advisory spans gateway, AutoRAG, EvalHub and pipeline images, regression-test RAG and serving workflows after the upgrade. Check which listed CVEs affect components the customer actually runs, and record compensating controls if an upgrade must wait.","delivery":"Delivery framing: put OpenShift AI on the same patch calendar as other critical infrastructure. Build a repeatable runbook covering image mirroring, operator upgrade, smoke tests for model serving and RAG, and rollback. Track errata like RHSA-2026:86891 in the vulnerability-management system with an owner and due date. Report patch latency to governance. Small IT teams should decide up front whether they will run these upgrades themselves or rely on a partner or managed service."},"roleDetails":{"sales":{"customerProblem":"Agencies adopting self-managed AI platforms underestimate the ongoing patching and upgrade workload.","stakeholders":["IT operations manager","CISO / vulnerability management","Platform engineering lead","Procurement / contract manager"],"discoveryQuestions":["Which OpenShift AI version and update channel are you on?","What is your patch SLA for Critical advisories on platform software?","Do you have staff experienced with OpenShift operator upgrades?","Are your clusters disconnected, requiring image mirroring?"],"valueHypothesis":"A defined patch runbook or managed operations can shorten time-to-patch for Critical AI platform advisories.","potentialEngagement":"OpenShift AI upgrade readiness review and patch runbook for one environment.","unsupportedClaims":["That the listed CVEs are exploitable in a given deployment","That upgrading resolves all AI-specific risks","Any specific time savings from managed services"]},"engineering":{"fit":"Applies to self-managed OpenShift AI deployments on the 3.5.x line.","architectureIntegration":"Operator-managed image updates through OpenShift; mirror registry for disconnected installs; regression tests for serving and RAG.","prerequisites":["Current subscription with errata access","Mirror registry for disconnected clusters","Test environment matching production"],"constraints":["CVE-level exploitability not assessed here","Upgrade may require maintenance windows","Disconnected sites need mirroring before upgrade"],"security":"Prioritize patching multi-tenant or broadly reachable AI gateway and RAG components; document compensating controls if patching is delayed.","proofOfValue":["Upgrade a test cluster to 3.5.2 and pass serving and RAG smoke tests","Measure time from advisory to production patch","Confirm image digests match the advisory after upgrade"]},"delivery":{"work":"Create and execute an OpenShift AI patch runbook including mirroring, upgrade, testing and rollback.","dependencies":["Maintenance window approval","Mirror registry capacity","Application owner test sign-off"],"ownership":"IT operations / platform engineering, with vulnerability management tracking.","skillsAdoption":"Operator upgrade and image mirroring training for operations staff.","governanceCheckpoints":["Change advisory board approval for upgrade","Post-upgrade security verification","Monthly patch latency report"],"acceptanceCriteria":["Proposed: Critical OpenShift AI advisories applied in production within the agency's Critical SLA","Proposed: rollback tested before production upgrade","Proposed: serving and RAG smoke tests pass after upgrade"],"risks":["Upgrade breaks custom pipelines or workbenches","Patch delays due to staffing","Disconnected sites lag behind connected ones"]}},"retrievedAt":"2026-10-30T03:09:23Z","enrichedAt":"2026-10-30T03:21:15Z","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Teams need OpenShift operator upgrade skills; small SLED IT shops may need managed services or partner support.","procurementImplications":"Subscription and support terms should include errata access and upgrade assistance. Managed-service contracts should specify patch SLAs for AI platform components.","operatingModelImplications":"Self-managed AI platforms add a recurring operations burden comparable to other Kubernetes platforms. Budget staff time for it.","sourceVerification":{"openedUrl":"https://access.redhat.com/errata/RHSA-2026:86891","referenceExcerpt":"Updated images are now available for Red Hat OpenShift AI.","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}