AI agents will only be useful in the enterprise if organizations can trust what they access, what they do and how they act on behalf of users.

This keynote looks at the foundation needed to make that possible: identity, access, data governance and Microsoft Purview.

The focus is practical: which permissions an agent uses, which business data it can reach, how sensitive information is protected and how actions remain visible and auditable.

A clear keynote on the next step of enterprise AI: not more hype, but control, transparency and trust.