CVE-2025-23245
First published: Thu May 01 2025(Updated: )
NVIDIA TensorRT-LLM for any platform contains a vulnerability in python executor where an attacker may cause a data validation issue by local access to the TRTLLM server. A successful exploit of this vulnerability may lead to code execution, information disclosure and data tampering.
Credit: psirt@nvidia.com
| Affected Software | Affected Version | How to fix |
|---|---|---|
| NVIDIA TensorRT-LLM | ||
| NVIDIA vGPU software |
Never miss a vulnerability like this again
Sign up to SecAlerts for real-time vulnerability data matched to your software, aggregated from hundreds of sources.
Reference Links
Frequently Asked Questions
What is the severity of CVE-2025-23245?
CVE-2025-23245 is considered to have a high severity due to the potential for code execution and data tampering.
How do I fix CVE-2025-23245?
To fix CVE-2025-23245, update to the latest version of NVIDIA TensorRT-LLM as recommended by NVIDIA.
What systems are affected by CVE-2025-23245?
CVE-2025-23245 affects the NVIDIA TensorRT-LLM on any platform where it is deployed.
What types of attacks can CVE-2025-23245 facilitate?
CVE-2025-23245 can facilitate attacks that lead to code execution, information disclosure, and data tampering.
Who is vulnerable to CVE-2025-23245?
Any user with local access to the TRTLLM server could be vulnerable to the exploitation of CVE-2025-23245.
- collector/mitre-cve
- source/MITRE
- agent/type
- agent/first-publish-date
- agent/source
- agent/author
- agent/event
- collector/nvd-api
- source/NVD
- agent/last-modified-date
- agent/severity
- agent/references
- agent/weakness
- agent/description
- agent/softwarecombine
- agent/tags
- agent/guess-ai
- agent/software-canonical-lookup
- vendor/nvidia
- canonical/nvidia tensorrt-llm
- canonical/nvidia vgpu software