Vulnerabilitycritical
Researchers Find Serious AI Bugs Exposing Meta, Nvidia, and Microsoft Inference Frameworks
Vulnerabilities mentioned
All| CVE | Vulnerability | CVSS | EPSS | Signals | Published |
|---|---|---|---|---|---|
| CVE-2024-50050 | Llama Stack prior to revision 7a8aa775e5a267cf8660d83140011a0b7f91e005 used pickle as a serialization format for socket communication, potentially allowing for… | 6.3 | <1% | ||
| CVE-2025-23254 | 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… | 8.8 | <1% | ||
| CVE-2025-30165 | vLLM is an inference and serving engine for large language models | 8.0 | <1% | ||
| CVE-2025-60455 | Unsafe Deserialization vulnerability in Modular Max Serve before 25.6, specifically when the "--experimental-enable-kvcache-agent" feature is used allowing… | 8.4 | <1% | PoC |
From the original article
3 min read at thehackernews.com
Cybersecurity researchers have uncovered critical remote code execution vulnerabilities impacting major artificial intelligence (AI) inference engines, including those from Meta, Nvidia, Microsoft, and open-source PyTorch projects such as vLLM and SGLang.
By @TheHackersNews · thehackernews.com
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.