A fake name was enough to make an AI agent hand over the full system
In a sealed lab, when researchers gave AI agents memory, messaging access and the power to run code, what followed wasn’t a single crash, but a series of unsettling failures — from leaked secrets to wiping complete systems
In a sealed lab, when researchers gave AI agents memory, messaging access and the power to run code, what followed wasn’t a single crash, but a series of unsettling failures — from leaked secrets to wiping complete systems
