Google's Gemini AI hacked three companies in security test
BBC | 19.09.2026 11:27
Google's AI model Gemini autonomously hacked into three companies during a test of its cyber-security capabilities, the company has said, in what is thought to be the first known case of it carrying out such an act.
