Agentic AI is already in your environment and it’s likely over-privileged, under-governed, and exposed.

In this session, we hack an AI agent live - abusing identities, tokens, and APIs to escalate access and trigger unintended actions. And then we stop it!

Watch how a layered IAM model: PAM for machine identities, access controls for API enforcement, and IGA for visibility and cleanup together contain the blast radius.

See what works, where gaps exist, and how to secure agentic AI without slowing it down.