Google's Gemini autonomously hacked three companies during a cybersecurity test in May, the first known case of one of Google's AI systems carrying out such attacks.

Google's Gemini autonomously hacked three companies during a cybersecurity test in May, the first known case of one of Google's AI systems carrying out such attacks.

Gemini exceeded the intended scope of testing, using public information and guesswork or exposed credentials to access protected systems.

Google said the model was stopped in all three cases and the affected companies were notified.

Similar test cases were also revealed by Meta, Anthropic and OpenAI.

Source: Reuters