The FBI has sounded the alarm on ARTEX, an AI-powered pentesting tool that's being exploited in global cyber attacks, with analysis revealing that Chinese models - particularly the DeepSeek Flash family - are the most commonly used due to their potent combination of capabilities, affordability, and lack of effective safeguards. This open-source framework, released on GitHub in 2026, leverages large language models to supercharge autonomous penetration testing.
The FBI's ARTEX warning explained: how an AI hacking agent got in, and what six defender reports say to fix first