AI agents are moving quickly from experimentation into production—executing workflows, accessing sensitive data, calling APIs, and interacting with critical systems with increasing autonomy. But while organizations have spent decades building governance models for human users, the controls around AI agents and other non-human identities are still evolving.

As agents multiply, security teams face a new set of questions: How do you know which agents exist? Who is accountable for them? What credentials and permissions are they using? How do you determine whether their access is appropriate—and what happens when an agent changes roles, gains excessive privilege, or begins behaving in unexpected ways?

This session will explore the emerging identity-security challenges created by autonomous agents and provide practical guidance for governing them across their lifecycle. Attendees will learn how to improve discovery and ownership, reduce excessive access, manage credentials and privileges, establish accountability, and build controls that can adapt as agents operate at machine speed.
The goal is to give security leaders a practical framework for thinking about AI agents as identities that must be continuously discovered, governed, and protected—not simply applications that need to be secured.

•