What is the CVSS score of CVE-2026-15467?

CVE-2026-15467 has a CVSS 3.1 base score of 8.1 (High). CVSS vector: CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:N. EPSS exploitation probability: 0.4%.

Which versions of OpenShift AI are affected by CVE-2026-15467?

Red Hat OpenShift AI's trustyai-service-operator contains an arbitrary code execution vulnerability (CVE-2026-15467, CVSS 8.1) allowing authenticated cluster users to bypass pod security policies,…. A patch is available.

How to fix or mitigate CVE-2026-15467?

Within 24 hours: inventory all Red Hat OpenShift AI deployments with trustyai-service-operator enabled, verify operator versions, and immediately restrict RBAC permissions for LMEvalJob creation to essential service accounts only; enable audit logging for all LMEvalJob resource modifications. Within 7 days: implement pod security standards (PSS) enforcement and network policies to isolate trustyai-service-operator, deploy admission controllers to validate sidecar configurations, and contact Red Hat Support to confirm patch timeline and request interim guidance. Within 30 days: apply the vendor-released patch immediately upon availability; if no patch timeline is established within 30 days, execute remediation alternatives such as decommissioning affected RHOAI components or migrating to alternative platforms depending on business criticality.

OpenShift AI CVE-2026-15467

EUVD-2026-55809 HIGH

Incorrect Privilege Assignment (CWE-266)

2026-08-10 redhat

GHSA-qw34-v3jw-3crv

RCE Red Hat Openshift Ai Rhoai Red Hat

8.1

CVSS 3.1 · Vendor: redhat

Severity by source

Vendor (redhat) PRIMARY

8.1 HIGH

AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:N

vuln.today AI

8.1 HIGH

Network-accessible via Kubernetes API with low privileges required; scope unchanged as execution is pod-local; no availability impact described.

3.1 AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:N

4.0 AV:N/AC:L/AT:N/PR:L/UI:N/VC:H/VI:H/VA:N/SC:N/SI:N/SA:N

Red Hat

8.1 HIGH

qualitative

Primary rating from Vendor (redhat).

CVSS VectorVendor: redhat

Attack Vector

Network

Attack Complexity

Low

Privileges Required

Low

User Interaction

None

Scope

Unchanged

Confidentiality

High

Integrity

High

Availability

None

Lifecycle Timeline

Analysis Generated

Aug 10, 2026 - 21:33 vuln.today

CVE Published

Aug 10, 2026 - 20:44 nvd

HIGH 8.1

DescriptionCVE.org

A flaw was found in the trustyai-service-operator's LMEvalJob controller. An authenticated user within the cluster can exploit this vulnerability by configuring a sidecar container to bypass existing security policies. This allows the user to enable and execute untrusted remote code, leading to arbitrary code execution within the cluster.

AnalysisAI

Arbitrary code execution in Red Hat OpenShift AI's trustyai-service-operator enables authenticated cluster users to bypass pod security policies via crafted sidecar container configurations in LMEvalJob resources. The CVSS 8.1 (High) score reflects network-accessible exploitation requiring only low-privilege cluster authentication, with high confidentiality and integrity impact against the affected workload context. No public exploit code has been identified and no CISA KEV listing exists at time of analysis, but the low barrier of entry (any authenticated cluster user) elevates insider-threat and compromised-account risk for RHOAI deployments.

Technical ContextAI

The trustyai-service-operator is a Kubernetes operator within Red Hat OpenShift AI (RHOAI) that manages LMEvalJob custom resources for running AI model evaluation workloads. CWE-266 (Incorrect Privilege Assignment) identifies the root cause: the LMEvalJob controller fails to adequately validate or restrict sidecar container specifications submitted by users, allowing them to inject containers with escalated capabilities or security context settings that circumvent admission controls or pod security standards. In Kubernetes/OpenShift, sidecar containers share the pod's network and optionally process namespace, making them a vector for security boundary violations. The affected CPE - cpe:2.3:a:red_hat:red_hat_openshift_ai_(rhoai):*:*:*:*:*:* - uses a wildcard version, indicating the issue is not yet bounded to a specific release range per NVD data.

RemediationAI

The primary remediation is to apply the vendor patch from Red Hat once released; consult https://access.redhat.com/security/cve/CVE-2026-15467 for the official advisory and exact fixed version, as no specific patched release version was confirmed in the available data. As an interim compensating control, restrict RBAC permissions to create, update, or patch LMEvalJob resources to only highly trusted users or service accounts - this does not eliminate the flaw but limits the population of potential attackers. Additionally, enforce OpenShift Pod Security Admission or a policy engine (e.g., Kyverno, OPA/Gatekeeper) to block sidecar container configurations that request elevated security contexts or privileged capabilities; note that this may require custom policy authoring and could interfere with legitimate evaluation jobs if policies are overly broad. Monitor the trustyai-service-operator namespace for anomalous container image pulls or outbound network connections from LMEvalJob pods as a detection measure.