Microsoft Events presents an Ignite session led by Kim Kischel, Joanne Marone, Irina Nechaeva, and Shilpa Ranganathan, spotlighting security and governance strategies for AI agents with Microsoft Agent 365.

Explore Microsoft Agent 365 Security and Governance Capabilities

Speakers: Kim Kischel, Joanne Marone, Irina Nechaeva, Shilpa Ranganathan
Session: BRK269 | Microsoft Ignite 2025 | Foundational (Level 100)

Overview

This Microsoft Ignite session introduces practical solutions for securing and governing an organization’s AI agent estate via Microsoft Agent 365. It covers:

The rise of assistive and autonomous AI agents, and associated security concerns
Strategies to control agent access using conditional access policies
Addressing new challenges linked to agent sprawl, vulnerability exposure, and data oversharing

Key Topics

AI Agent Security & Oversight

Enforcing comprehensive governance to prevent data leaks and oversharing
Utilizing security policies to tightly regulate agent permissions
Traceability and monitoring with audit trails

Conditional Access Controls

Implementing conditional access to restrict agent operations
Reviewing agent activity logs and access events

Data Loss Prevention (DLP)

DLP matching and detection for sensitive activity
Automated alerts and response to risky agent behaviors

Compliance & Audit

Utilizing Purview Audit for oversight of all agent interactions
Ensuring regulatory compliance through systematic event logging

Threat Response & Investigation

Reviewing attack path visualizations and exposure risks
Investigating incidents involving agents—case studies of collection incidents
Formulating security recommendations and effective remediation steps

Governance Best Practices

Regular review of agent sprawl and permission boundaries
Close coordination between admins and security teams
Proactive threat hunting and incident management procedures

Additional Resources

For further details, explore on-demand sessions and governance resources provided by Microsoft Events.