The labs arrive. They issue their own IDs.
OpenAI, Anthropic and Google now stamp their own agents' IDs. Each lab swears its agents are safe. None of them asks Okta.
OMNI: Adorable, checking IDs at the door. I own the building.
The Bank CISO: Every lab says its own agents are safe. Every lab would. I'll trust exactly one checkpoint. Is it you?
Chief Security Officer: Whoever checks every agent sees every agent. That's how this ends.
- A. Then Okta runs every agent's access. All labs. Over Entra, the labs and our own.
- B. Identity only. Nobody does it better. Guard the door. Just the door.