CVE-2025-12805: Improper Isolation or Compartmentalization in Red Hat Red Hat OpenShift AI 2.25
A flaw was found in Red Hat OpenShift AI (RHOAI) llama-stack-operator. This vulnerability allows unauthorized access to Llama Stack services deployed in other namespaces via direct network requests, because no NetworkPolicy restricts access to the llama-stack service endpoint. As a result, a user in one namespace can access another user’s Llama Stack instance and potentially view or manipulate sensitive data.
AI Analysis
Technical Summary
The vulnerability CVE-2025-12805 in Red Hat OpenShift AI 2.25 involves improper isolation or compartmentalization within the llama-stack-operator component. Specifically, the absence of NetworkPolicy restrictions on the llama-stack service endpoint permits users in one Kubernetes namespace to send direct network requests to Llama Stack services in other namespaces. This lack of network segmentation breaks namespace isolation, potentially exposing sensitive data or allowing unauthorized manipulation of another user's Llama Stack instance. The CVSS v3.1 score is 8.1 (high severity), reflecting network attack vector, low attack complexity, required privileges, no user interaction, and high confidentiality and integrity impacts. Red Hat has released updated container images in RHOAI 2.25.2 and advises upgrading clusters following their documentation. The vendor advisory notes no direct code fixes but implies mitigation through updated images and configuration guidance.
Potential Impact
Successful exploitation allows an attacker with privileges in one Kubernetes namespace to bypass network isolation controls and access Llama Stack services in other namespaces. This can lead to unauthorized disclosure and modification of sensitive data managed by those services. The vulnerability does not affect availability. There are no known public exploits currently. The impact is rated high due to the potential confidentiality and integrity breaches across namespace boundaries in a multi-tenant environment.
Mitigation Recommendations
Red Hat has released updated container images as part of RHOAI 2.25.2. While no direct code fixes are included, users should upgrade to this version and carefully follow Red Hat's official upgrade and configuration instructions to fully apply the errata and mitigate the vulnerability. Applying appropriate NetworkPolicy configurations to restrict cross-namespace access to the llama-stack service endpoint is critical. Patch status is confirmed via the vendor advisory. No additional urgent actions are indicated beyond upgrading and applying the recommended configuration changes.
CVE-2025-12805: Improper Isolation or Compartmentalization in Red Hat Red Hat OpenShift AI 2.25
Description
A flaw was found in Red Hat OpenShift AI (RHOAI) llama-stack-operator. This vulnerability allows unauthorized access to Llama Stack services deployed in other namespaces via direct network requests, because no NetworkPolicy restricts access to the llama-stack service endpoint. As a result, a user in one namespace can access another user’s Llama Stack instance and potentially view or manipulate sensitive data.
AI-Powered Analysis
Machine-generated threat intelligence
Technical Analysis
The vulnerability CVE-2025-12805 in Red Hat OpenShift AI 2.25 involves improper isolation or compartmentalization within the llama-stack-operator component. Specifically, the absence of NetworkPolicy restrictions on the llama-stack service endpoint permits users in one Kubernetes namespace to send direct network requests to Llama Stack services in other namespaces. This lack of network segmentation breaks namespace isolation, potentially exposing sensitive data or allowing unauthorized manipulation of another user's Llama Stack instance. The CVSS v3.1 score is 8.1 (high severity), reflecting network attack vector, low attack complexity, required privileges, no user interaction, and high confidentiality and integrity impacts. Red Hat has released updated container images in RHOAI 2.25.2 and advises upgrading clusters following their documentation. The vendor advisory notes no direct code fixes but implies mitigation through updated images and configuration guidance.
Potential Impact
Successful exploitation allows an attacker with privileges in one Kubernetes namespace to bypass network isolation controls and access Llama Stack services in other namespaces. This can lead to unauthorized disclosure and modification of sensitive data managed by those services. The vulnerability does not affect availability. There are no known public exploits currently. The impact is rated high due to the potential confidentiality and integrity breaches across namespace boundaries in a multi-tenant environment.
Mitigation Recommendations
Red Hat has released updated container images as part of RHOAI 2.25.2. While no direct code fixes are included, users should upgrade to this version and carefully follow Red Hat's official upgrade and configuration instructions to fully apply the errata and mitigate the vulnerability. Applying appropriate NetworkPolicy configurations to restrict cross-namespace access to the llama-stack service endpoint is critical. Patch status is confirmed via the vendor advisory. No additional urgent actions are indicated beyond upgrading and applying the recommended configuration changes.
Technical Details
- Data Version
- 5.2
- Assigner Short Name
- redhat
- Date Reserved
- 2025-11-06T13:48:05.305Z
- Cvss Version
- 3.1
- State
- PUBLISHED
Threat ID: 69c5ac523c064ed76fd41c25
Added to database: 3/26/2026, 9:59:46 PM
Last enriched: 4/3/2026, 1:03:49 PM
Last updated: 5/11/2026, 6:28:27 AM
Views: 52
Community Reviews
0 reviewsCrowdsource mitigation strategies, share intel context, and vote on the most helpful responses. Sign in to add your voice and help keep defenders ahead.
Want to contribute mitigation steps or threat intel context? Sign in or create an account to join the community discussion.
Actions
Updates to AI analysis require Pro Console access. Upgrade inside Console → Billing.
Need more coverage?
Upgrade to Pro Console for AI refresh and higher limits.
For incident response and remediation, OffSeq services can help resolve threats faster.
Latest Threats
Check if your credentials are on the dark web
Instant breach scanning across billions of leaked records. Free tier available.