Choose how you want to continue.
Critical Hugging Face Transformers flaw ran attacker code on a routine model load
News
Security
The-Latest
AI Security
Cybersecurity
Vulnerabilities
artificial-intelligence
Hugging Face
Summary
A critical vulnerability in Hugging Face's Transformers library, tracked as CVE-2026-4372, has been disclosed by Pluto Security Inc., allowing attacker-controlled AI models to execute arbitrary code on victim machines through standard model-loading commands, regardless of security measures advised by Hugging Face.
MORE POSTS LIKE THIS
New details on OpenAI/Hugging Face attack emerge as security industry debates AI agent controls
Siliconangle ·
Aug 6, 2026
Further Developments About Internal AI Models Hacking Things
Thezvi Substack ·
Aug 2, 2026
Even if 90% of AI-discovered vulnerabilities are made up, it's still terrifying
Guillaume Kerkour ·
Jul 13, 2026
Critical Squid Proxy Vulnerabilities Patched in Latest Release
Securityonline ·
Jun 13, 2026
Cisco’s Risk-Based Vulnerability Disclosure in the Age of AI
Blogs Cisco ·
May 22, 2026
AUTHOR