We’ve Been Securing AI Agents Wrong

This video is sponsored by Teleport. Learn more about Teleport’s Agentic Identity Framework: https://fandf.co/4hyPyfG AI agents can now read your tickets, query your databases and change your infrastructure on their own, and most of them do it with a shared password. In this video I look at why agents need their own identity, then run the same task twice on a test store. With one shared admin login, the agent deleted more than a million audit-log rows, and the database log could not say who did it. With every agent getting its own identity through Teleport, the same model was refused and asked a human instead. Links: Teleport Agentic Identity Framework: https://fandf.co/4hyPyfG The lethal trifecta for AI agents (Simon Willison): https://simonwillison.net/2025/Jun/16/the-lethal-trifecta/