CVE-2026-17557 Red Hat CVE debrief

Who should care

Red Hat OpenShift AI administrators, security teams, and users with access to lm-evaluation-harness should assess exposure and implement security controls. This includes reviewing access controls, monitoring for suspicious activity, and implementing additional security measures to protect against potential exploitation. Operators, platform administrators, and vulnerability management teams should also be aware of the potential impacts and take necessary to

Why it matters

CVE-2026-17557 allows authenticated users to exploit lm-evaluation-harness in Red Hat OpenShift AI, potentially disclosing sensitive information and enabling SSRF attacks. Administrators and security teams should review access controls, monitor for suspicious activity, and implement additional security measures.

• Disclosure of tenant model authentication secrets to attacker-controlled external listeners
• SSRF attacks against internal network endpoints reachable from the evaluation job pod
• Potential for lateral movement within the OpenShift AI environment
• Need for verification of affected versions and remediation steps

Technical summary

The lm-evaluation-harness component of Red Hat OpenShift AI is vulnerable to Server-Side Request Forgery (SSRF) attacks and disclosure of tenant model authentication secrets. An authenticated user can submit a malicious model URL in the JobSpec to exploit this vulnerability, potentially disclosing sensitive information and enabling SSRF attacks against internal network endpoints reachable from the evaluation job pod. This vulnerability is a result of insufficient validation and sanitization of user-supplied input in the JobSpec model URL.

Defensive priority

Medium priority for Red Hat OpenShift AI administrators and security teams

Recommended defensive actions

• Review and restrict access to lm-evaluation-harness in Red Hat OpenShift AI
• Monitor for suspicious evaluation job submissions
• Implement additional security controls for model URLs and authentication secrets
• Verify affected versions and apply remediation steps
• Conduct regular security audits and vulnerability assessments
• Enhance monitoring and detection capabilities for exposed assets
• Track exceptions and retest remediated assets to ensure thorough mitigation

Evidence notes

The CVE record and source item provide details on the vulnerability in lm-evaluation-harness, allowing for secret disclosure and SSRF attacks. Red Hat OpenShift AI is affected, but specific version information is not provided.

Sources and references

Methodology and review provenance

AI-assisted synthesis based on stored public vulnerability evidence. System validation, approval state, and publication status do not by themselves establish human review of this revision. PatchSiren helps prioritize defensive review and does not prove exposure or remediation on any system.