An unauthenticated attacker can trigger a denial of service condition in the NVIDIA Triton Inference Server for Linux by exploiting improper input validation.

This vulnerability is caused by improper validation of specified quantities in input, which occurs during the processing of network requests. As the CVSS vector indicates an attack vector of network (AV:N) and no required privileges (PR:N), the vulnerability is exploitable by unauthenticated remote attackers.

The exploitation of this vulnerability leads to a denial of service, which can render the inference server unavailable for critical machine learning and artificial intelligence workloads. Given the CVSS score of 7.5, this high severity flaw poses a significant risk to operational continuity, potentially disrupting automated decision-making processes that rely on the Triton server.

Immediate Action: Update the NVIDIA Triton Inference Server to release r25.10 or later to ensure the input validation logic is corrected.

Proactive Monitoring: Monitor server logs for unexpected process crashes, sudden spikes in resource consumption, or irregular input patterns that deviate from expected API request schemas.

Compensating Controls: Deploy a Web Application Firewall or an API gateway to enforce strict schema validation on incoming requests, which can help block malformed inputs before they reach the server.

Public Exploit Available: No (exploit_available: unknown)

Organizations utilizing NVIDIA Triton Inference Server should prioritize the update to version r25.10 as part of their next maintenance cycle. Because this vulnerability allows for unauthenticated denial of service, leaving the server exposed to untrusted networks significantly increases the risk of service disruption.