The vulnerability exists in the odh-model-controller component of Red Hat OpenShift AI (RHOAI). The loadSecret function improperly reads the Secret namespace from user-controlled input without validation, allowing an authenticated user with permissions to create custom resources to read sensitive secrets from other namespaces. This 'confused deputy' flaw enables cross-namespace secret disclosure, including high-value Red Hat OpenShift AI-managed secrets. The CVSS v3.1 score is 6.5 (medium severity) with network attack vector, low attack complexity, low privileges required, no user interaction, unchanged scope, and high confidentiality impact. No integrity or availability impact is noted. There is no explicit patch or remediation level stated in the vendor advisory as of the publication date.

An attacker with authenticated access and permissions to create custom resources can read sensitive API keys and cloud credentials from namespaces they should not have access to. This leads to information disclosure of potentially high-value secrets managed by Red Hat OpenShift AI. There is no indication of integrity or availability impact. No known exploits in the wild have been reported.

Patch status is not yet confirmed — check the Red Hat advisory at https://access.redhat.com/security/cve/CVE-2026-16456 for current remediation guidance. The advisory does not currently state that a fix is available or provide a remediation level. Until a fix is released, restrict permissions to create custom resources to trusted users only to reduce risk. Monitor Red Hat's updates for any forthcoming patches or mitigations.