Red Hat Red Hat OpenShift AI (RHOAI)

• CVE-2026-83495 Unrated Trainer: trainer: trainjob env passthrough enables valuefrom.secretkeyref injection for namespace secret exfiltration –
• CVE-2026-107912 High 7.7 Io.quarkus/quarkus-vertx-http: quarkus-vertx-http: authorization bypass in management interface via path concatenation flaw 0.2%
• CVE-2026-117574 Critical 9.1 Ai-gateway-payload-processing: ai-gateway-payload-processing: tenant/credential selection from client-controlled url path with no in-process authorization 0.2%
• CVE-2026-13451 High 7.7 Kserve: kserve: server-side request forgery via malicious storageuri 0.2%
• CVE-2026-114236 Low 3.3 Podman: goroutine/file-descriptor leak in storage/pkg/archive.applylayer and applyuncompressedlayer on malformed layers 0.1%
• CVE-2026-17557 Medium 6.5 Lm-evaluation-harness: lm-evaluation-harness: ssrf via job-spec model.url with bearer credential forwarding 0.3%
• CVE-2026-108334 High 7.4 Github.com/kuadrant/kuadrant-operator/internal/wasm: github.com/rh-api-management/rhcl-operator-product-build/kuadrant-operator: kuadrant-operator: crafted httproute match values can bypass gateway policy overrides 0.3%
• CVE-2026-102010 High 7.0 Gcc-toolset-15-gcc: gcc: gcc-toolset-16: gcc: denial of service via use-after-free in binary heap erase_if 0.3%
• CVE-2026-96889 High 7.8 Librsvg: use-after-free when xml includes have duplicated entities 0.6%
• CVE-2026-87743 High 7.5 Quarkus-vertx-http: authorization bypass via path normalization discrepancy in quarkus http security 0.5%
• CVE-2026-92091 Medium 5.9 Jwcrypto: jwcrypto: denial of service via o(n^2) duplicate check on unbounded jwk key_ops array 0.6%
• CVE-2026-86332 Medium 6.5 Odh-dashboard: odh-dashboard: nim credential secret readable by any authenticated user 0.4%
• CVE-2026-84185 Medium 5.9 Jwcrypto: jwcrypto: general json jws kid binding bypass during jwkset verification 0.1%
• CVE-2026-18393 Medium 5.4 Ffmpeg: ffmpeg: heap buffer overflow in tdsc_load_cursor() via cur_fmt_mono cursor 0.4%
• CVE-2026-80179 Medium 5.9 Jwcrypto: jwcrypto: denial of service via malformed jwe tokens 0.3%
• CVE-2026-18950 High 8.8 Odh-dashboard: odh-dashboard: confused-deputy privilege escalation via unchecked roleref in rolebinding creation 0.5%
• CVE-2026-18948 Critical 9.9 Feast: feast: unsafe dill deserialization of registry-stored udfs — rce on feature server and registry server 1.1%
• CVE-2026-18947 High 8.5 Feast: feast: authorization bypass in /materialize endpoints enables dos via unauthorized full re-materialization 0.6%
• CVE-2026-18942 Medium 5.5 Feast-operator: feast: feast apply cronjob runs user python with feature-server sa — tenant code to sa token escalation 0.4%
• CVE-2026-18941 High 7.7 Feast: feast-operator: feast: default authentication mode is no_auth — shared multi-tenant instances deployed without authentication 0.9%
• CVE-2026-14450 Critical 9.9 Maas-billing: maas api: privilege escalation via forged http headers due to missing authentication 0.7%
• CVE-2026-13717 High 8.8 Rhoai maas: llm-d: maas/llm-d inference gateway: default allowedroutes.namespaces.from: all allows namespace users to hijack shared model-serving traffic (tokens, prompts, outputs) 0.6%
• CVE-2026-23538 High 7.5 Feast: resource exhaustion via websocket endpoint 0.9%
• CVE-2026-15063 Medium 6.3 Trustyai-service-operator: trustyai service operator: gorch port bypass when auth is enabled 0.3%
• CVE-2026-15044 Medium 6.3 Trustyai-service-operator: trustyai service operator: unauthenticated access to ai guardrails and orchestrator apis 0.2%
• CVE-2026-23537 Critical 9.1 Feast: unauthenticated arbitrary file write 1.0%
• CVE-2026-12706 Medium 6.5 Ffmpeg: ffmpeg: heap use-after-free read in rasc decoder decode_move() 0.5%
• CVE-2026-12491 Medium 4.8 Vllm: vllm: image exif rotation & png trns transparency not normalized, causing mismatch between model input and expectations 0.2%
• CVE-2026-11850 Medium 5.0 Krb5: krb5: integer underflow in berval2tl_data() leads to heap out-of-bounds read 0.3%
• CVE-2026-6385 Medium 6.5 Ffmpeg: ffmpeg: denial of service and potential arbitrary code execution via signed integer overflow in dvd subtitle parser 0.4%
• CVE-2026-5483 High 8.5 Odh-dashboard: odh dashboard kubernetes service account exposure 0.5%
• CVE-2025-12805 High 8.1 Llama-stack-k8s-operator: llama stack service exposed across namespaces due to missing networkpolicy 0.4%
• CVE-2026-23536 High 7.5 Feast: unauthenticated arbitrary file read 2.5%
• CVE-2025-13327 Medium 6.3 Uv: uv: specially crafted zip archives lead to arbitrary code execution due to parsing differentials 0.2%
• CVE-2025-11065 Medium 5.3 Github.com/go-viper/mapstructure/v2: go-viper's mapstructure may leak sensitive information in logs in github.com/go-viper/mapstructure 0.5%
• CVE-2026-0603 High 8.3 Org.hibernate/hibernate-core: hibernate: information disclosure and data deletion via second-order sql injection 1.0%
• CVE-2025-12103 Medium 5.0 Openshift-ai: trusty ai grants all authenticated users to list pods in any namespace 0.3%
• CVE-2025-6193 Medium 5.9 Trustyai-explainability: command injection via lmevaljob cr 0.7%
• CVE-2025-4574 Medium 6.5 Crossbeam-channel: crossbeam-channel vulnerable to double free on drop 0.6%