CVE-2025-12103

The record

A flaw was found in Red Hat Openshift AI Service. The TrustyAI component is granting all service accounts and users on a cluster permissions to get, list, watch any pod in any namespace on the cluster. TrustyAI is creating a role `trustyai-service-operator-lmeval-user-role` and a CRB `trustyai-service-operator-default-lmeval-user-rolebinding` which is being applied to `system:authenticated` making it so that every single user or service account can get a list of pods running in any namespace on the cluster Additionally users can access all `persistentvolumeclaims` and `lmevaljobs`

Affected packages

Related vulnerabilities

• CVE-2026-63349 AnyIO is a high level asynchronous concurrency and networking framework that works on top… CWE-266
• CVE-2026-62106 Subscriber Privilege Escalation in SMS Alert Order Notifications <= 3.9.9 versions. CWE-266
• CVE-2026-62102 Subscriber Privilege Escalation in Gato GraphQL <= 19.2.3 versions. CWE-266
• CVE-2026-8303 Incorrect privilege assignment vulnerability in TUBITAK BILGEM Software Technologies… CWE-266
• CVE-2026-81805 Unauthenticated Privilege Escalation in SiteSkite <= 2.1.5 versions. CWE-266
• CVE-2026-15140 A privilege-escalation issue in the Portworx Operator when deployed on Red Hat OpenShift… CWE-266
• CVE-2026-86804 A vulnerability was identified in seakee CPA-Manager-Plus up to 1.11.10. CWE-266
• CVE-2026-77654 Improper Privilege Management vulnerability in Horizon Security Analyzer (formerly… CWE-266