You describe a world of sprawling non-human identities, and we're hearing about that a How does that agentic AI make an identity problem that organizations may already have even harder to see and control?

So in practice, non-human identities, they've really been sprawling for a while.

This part isn't necessarily a new problem.

It's a pre-agentic problem that agents really blow up.

Because if we think about it, if you've been around for a while and you've got Directory service accounts, running databases, applications, you name it.

Every one of those credentials is something that can be potentially exploited.

Looking at a DevOps pipeline, you've got that pipeline's access control service principles, you name it, for performing cloud operations.

nothing absolutely nothing