How does your body fight off millions of threats it has never seen before, without a single rulebook?

That question might hold the key to securing AI agents.

Anders Åskåsen sits down with Tom from FSP to explore a fresh way of thinking about identity: borrow from biology.

They dig into why role-based access keeps exploding, why we never solved least privilege even for simple service accounts, and why agents now run those same old problems millions of times faster.

Tom's big idea? Treat security like an immune system. Block the known-bad on sight. Slow down the suspicious. Give a human time to step in.

Plus the wild story of an AI agent that signed up for its own API token overnight and would not stop making phone calls.

If you own identity, security, or governance, press play.