As AI agents move from answering questions to executing business actions, identity becomes a critical security boundary.
This session demonstrates how Microsoft Foundry Toolboxes and Microsoft Entra identity enable governed access to enterprise tools, APIs, and MCP servers. Using an employee-service scenario, we’ll compare user delegation, agent identity, and managed identity while showing how permissions, credentials, and consent remain isolated from agent logic.
We’ll cover least-privilege access, tool governance, auditability, human approval, and Responsible AI safeguards. Attendees will leave with a practical architecture for building agents that can act on behalf of users without compromising enterprise security.