CVE-2026-15154
MediumCVSS 6.5Exploitation Probability (EPSS)
Low risk38th percentile - higher than 38% of all known CVEs
Summary
A ReDoS vulnerability was found in the `guardrails-detectors` component of Red Hat OpenShift AI. A remote attacker can send crafted regular expressions to the public detection API, causing catastrophic backtracking and 100% CPU usage by a worker process.
Risk Assessment
This attack results in a denial of service (DoS) for the entire guardrails-mediated LLM pipeline, potentially halting AI applications and impacting critical service availability.
Recommendation
Immediately update the `guardrails-detectors` component to a patched version and consider restricting access to the public detection API to trusted IP addresses only.
Other vulnerabilities in guardrails-detectors
See all- CVE-2026-15143Critical
A flaw was found in the file_type content detector of guardrails-detectors, allowing a remote attacker to supply an arbitrary XSD string processed without restrictions. This can lead to server-side requests to arbitrary URLs or local file reads.
- CVE-2026-15378Critical
A flaw in the `guardrails-detectors` component allows a remote attacker to perform a blind SSRF attack by submitting a crafted XSD string. This can lead to unauthorized access to sensitive information, including credentials from cloud metadata services, Kubernetes API, internal MinIO, and other internal network endpoints. It also enables local file reads of critical data such as service account tokens and pod secrets.
Original NVD description (English source)
A flaw was found in `guardrails-detectors`, a component of Red Hat OpenShift AI. This vulnerability, known as Regular Expression Denial of Service (ReDoS), allows a remote attacker to provide specially crafted regular expressions to the public detection API. This can cause catastrophic backtracking, leading to a worker process consuming 100% CPU indefinitely and resulting in a denial of service for the entire guardrails-mediated LLM pipeline.