Microsoft Agent 365 Security and Governance Capabilities - BRK269

Discover comprehensive security and governance capabilities for AI agents through this 41-minute conference session from Microsoft Ignite 2025. Learn how to confidently activate your AI agent estate using Microsoft Agent 365, whether deploying assistive agents or fully autonomous digital workers. Explore essential strategies for managing AI access and preventing agent sprawl while maintaining strict control and oversight over your AI infrastructure. Master techniques for preventing data oversharing and leaks through conditional access policies and data loss prevention measures. Understand how to defend against emerging AI threats and vulnerabilities while ensuring compliance with AI regulations through comprehensive audit capabilities. Examine real-world scenarios including incident investigation procedures, security recommendations, and remediation steps for AI agents. Analyze attack path visualization techniques and learn to identify potential lateral movement risks in AI-enabled environments. Gain insights into event traceability, regulatory compliance through Purview Audit, and administrative review processes for agent activity monitoring.

Syllabus

0:00 - Irina introduces co-speakers Shilpa and Kim and contextualizes AI agents trend
00:00:52 - Security concerns with AI agents and need for control and oversight
00:13:02 - Protecting agent access with conditional access policies
00:18:10 - Overview of new security challenges with AI agents
00:23:52 - Admin reviews agent activity, DLP matches, and event traceability
00:25:42 - Purview Audit captures all agent interactions ensuring regulatory compliance
00:31:11 - Investigation begins on low-priority collection incident involving an agent
00:35:16 - Exploring security recommendations and remediation steps for AI agents
00:36:21 - Reviewing attack path visualization and potential lateral movement risks