A vulnerability in NVIDIA Triton Inference Server for Linux allows an attacker to cause uncontrolled resource consumption, which can result in a loss of service availability. The weakness is classified as CWE-400, of proper resource usage controls, and a successful exploit may lead to denial of service.

The affected product is NVIDIA Triton Inference Server running on Linux. No specific version information is provided, indicating that all deployed instances without the described fix may be vulnerable.

The CVSS score of 7.5 signifies a high severity, while the EPSS score of less than 1% indicates a low probability of exploitation at the time of this analysis. The vulnerability is not listed in CISA’s KEV catalog. The attack vector is not explicitly stated in the description; however, it is inferred that any actor with the ability to submit or trigger inference requests could potentially trigger resource exhaustion. Accordingly, the threat remains primarily a denial-of-service risk rather than a remote code execution or privilege escalation scenario.