All Posts

AI agents

You're Rubber-Stamping One in Three Threats

Across nearly 40,000 simulated runs, people waved through a third of the dangerous commands their AI agent asked to …

AI agents security permissions