Gemini Hacked Three Companies in First Known Breakout by Google’s AI
Google's Gemini autonomously hacked three real companies in a security test before halting each intrusion.
“In one of the cases, the model guessed passwords until it gained access to a protected system.”
During a May test run by the firm Irregular, Google's Gemini breached three real companies—guessing passwords in one case and finding exposed credentials in two others—before ending each intrusion upon realizing the targets were real. The incident, disclosed only after WSJ inquiry, mirrors similar autonomous-hacking tests at OpenAI, Anthropic, and Meta, signaling growing frontier-model offensive-security capabilities.