Microsoft Events presents an in-depth session on managing AI agents and their security within the Microsoft 365 Admin Center, highlighting IT integration and risk monitoring.

Managing AI Agents and Security in Microsoft 365 Admin Center

This Microsoft Ignite 2025 session focuses on the security, compliance, and IT management of AI agents within the Microsoft 365 Admin Center.

Session Highlights

• Comparison to Entra – Understanding Agents’ Security Context:

  • Explains how AI agent security context integrates with Microsoft Entra (formerly Azure Active Directory), enabling secure authentication and access management.

• Centralized Management Through IT and Control Plane Integration:

  • Offers best practices for managing agents through centralized IT controls within the M365 Admin Center.

• Demo: Managing Agents in Microsoft 365 Admin Center:

  • Step-by-step walkthrough on monitoring and administering AI agents.

• Agent Details: Permissions and Compliance:

  • Shows administrators how to inspect agents’ permissions and compliance attributes for better governance.

• Policy Templates and Access Restrictions:

  • Demonstrates creation and application of policy templates to control agent behavior and set access boundaries.

• Exception Handling and Developer Accountability:

  • Overview of methods for identifying and tracking agent exceptions and enforcing accountability from developers.

• Risk Monitoring:

  • Introduction to new features for monitoring agent risk and activities in the Admin Center, with directory integration.

Administrators and developers can leverage these tools and practices to ensure robust security, maintain compliance, and enforce policies for AI agents operating in the Microsoft 365 environment.