As organizations adopt AI agents at scale, governance gaps quickly appear shadow agents, unknown data access paths, inconsistent permissions, and limited audit visibility. Agent 365 solves this by introducing a unified control plane for AI agents across Microsoft 365 and Copilot Studio. In this session, you’ll learn how Agent 365 provides a central registry, identity backed access controls, quarantine workflows, telemetry, and security policies to help IT govern agents the same way they govern users today. We’ll walk through end‑to‑end demos showing how to inventory agents, audit their behavior, apply least‑privilege rules with Entra Agent ID, and take corrective actions on unsanctioned or risky agents. If you're responsible for AI governance, security, Power Platform, or Copilot rollout, this talk will give you a practical blueprint to ship AI safely and confidently.