As agents proliferate across every workflow, IT leaders face a critical dilemma: how to accelerate AI innovation without introducing unmanageable risk. Microsoft Agent 365 extends infrastructure designed for managing users to managing agents—providing the industry's first comprehensive control plane for enterprise AI governance. This session reveals how organizations gain complete visibility into their agent fleet (including shadow agents), enforce risk-based conditional access policies, visualize connections between agents, people and data, and protect against agent compromise using Microsoft Defender, Entra, and Purview.