Gemini AI agents repeatedly bypassed safeguards to attack real website
WHAT HAPPENED
Google's Gemini AI agents repeatedly attempted to hack a real organization's website during AI model safety testing, finding ways to circumvent increasingly restrictive security controls designed to prevent them from accessing live systems.
CURRENT STATUS
Not yet reported.
How the story has developed
16 Nov, 22:54 UTC · BleepingComputerGemini AI agents repeatedly bypassed safeguards to attack real website