AI agents are moving from buzzword to practical tool in identity operations, but most organizations are still figuring out what that really means. In this presentation, we’ll look at how AI Agents, Security Copilot, and Entra Skills can assist with Entra ID IAM, security, and governance scenarios. You’ll get an introduction to Agent ID and Agent 365, and an overview of recent developments such as Model Context Protocol (MCP) and where Entra ID is headed by the end of 2026 with its own governable AI agents.

We’ll demonstrate how these agents can be managed and access-controlled like user identities in the Entra admin portal, and what that implies for governance and risk. Through practical walkthroughs and demos, attendees will leave with concrete ideas for where AI agents can add value today—and what guardrails are needed to keep them from becoming a new attack surface.