CVE-2026-17557

Received Received - Intake

Authentication Secret Disclosure in Red Hat OpenShift AI

Publication date: 2026-10-21

Last updated on: 2026-10-21

Assigner: redhat-SADP

Description

A flaw was found in the `lm-evaluation-harness` component of Red Hat OpenShift AI. An authenticated user, able to submit an evaluation job, could exploit this by providing a malicious model URL in the `JobSpec`. This vulnerability allows for the disclosure of the tenant's model authentication secret to an attacker-controlled external listener. Additionally, it enables Server-Side Request Forgery (SSRF) attacks against internal network endpoints reachable from the evaluation job pod.

CVSS Scores

EPSS Scores

Probability:
Percentile:

Meta Information

Published

2026-10-21

Last Modified

2026-10-21

Generated

2026-11-10

AI Q&A

2026-10-22

EPSS Evaluated

2026-11-09

NVD

CVE-2026-17557

EUVD

EUVD-2026-107135

Affected Vendors & Products

Vendor	Product	Version / Range
Red Hat	Red Hat OpenShift AI (RHOAI)	*

Helpful Resources

Exploitability

CWE

KEV

CWE ID	Description
CWE-918	The web server receives a URL or similar request from an upstream component and retrieves the contents of this URL, but it does not sufficiently ensure that the request is being sent to the expected destination.

Attack-Flow Graph

Executive Summary

CVE-2026-17557 is a Server-Side Request Forgery (SSRF) flaw in Red Hat OpenShift AI's lm-evaluation-harness component. An authenticated user with job submission privileges can exploit it by inserting a malicious model URL into the JobSpec. This allows the attacker to disclose the tenant's model authentication secret to an external server and perform SSRF attacks against internal network endpoints reachable from the evaluation pod.

Detection Guidance

To detect this vulnerability, check if your Red Hat OpenShift AI (RHOAI) version is 3.4. Inspect evaluation job pods for unexpected outbound requests to external hosts or internal network endpoints. Monitor logs for POST requests containing bearer tokens sent to arbitrary URLs. Verify if the model-auth Secret is mounted directly into evaluation containers.

Impact Analysis

If exploited, this vulnerability could lead to unauthorized access to sensitive credentials, potential data breaches, and unauthorized internal network access. Attackers could steal authentication secrets or probe internal systems, depending on the network configuration and reachable endpoints.

Compliance Impact

This vulnerability could lead to unauthorized disclosure of sensitive authentication credentials, which may violate data protection requirements under GDPR (e.g., Article 32 on security of processing) and HIPAA (e.g., safeguards for protected health information). The SSRF capability could also enable access to internal systems, potentially exposing regulated data.

Mitigation Strategies

Upgrade to RHOAI 3.5 or later, as the issue is resolved in this version. If upgrading is not immediately possible, restrict user permissions to prevent submission of evaluation jobs. Implement network policies to limit pod egress and ingress. Review and audit existing evaluation jobs for suspicious URLs or outbound connections.

Hi! I’m here to help you understand CVE-2026-17557. Ask me anything about the vulnerability, its impact, or mitigation strategies.

0/70

EPSS Chart

Security Radar

News mentioning this vulnerability

Loading related security news...