Black Friday sale: an extra 40% off every course until Monday, November 30.New Year sale: an extra 30% off every course until Friday, January 1.Memorial Day sale: an extra 25% off every course until Monday, May 31.SysAdmin Day sale: an extra 25% off every course until Sunday, August 1.Back to School sale: an extra 25% off every course until Sunday, August 22.Labor Day sale: an extra 25% off every course until Monday, September 6.Black Friday sale: an extra 40% off every course until Monday, November 29.New Year sale: an extra 30% off every course until Saturday, January 1.Choose your certification →

Copyright (c) 2026 MindMesh Academy. All rights reserved. This content is proprietary and may not be reproduced or distributed without permission.

3.3.1. Monitoring Agent Activity

💡 First Principle: Monitoring agents means answering three questions — how much they're used, whether they're failing, and whether anything they do looks risky. Usage and health come from Agent 365 analytics; risk comes from the security products feeding into the same registry.

Agent overview (Microsoft 365 admin center > Agents > Overview) covers the last 30 days: the number of agents in the registry, active users, agent run-time hours and registry sync status, plus action cards for pending requests, agents at risk, agents without owners and agents with exceptions, and analytics by creator (your organization, third party, Microsoft), top platforms and trending agents.

Per-agent activity is on the agent details Activity tab: active users, sessions (a new session starts after 30 minutes of inactivity), exceptions and run-time, adjustable from the 30-day default. Activity metrics for third-party agents need E7 or Agent 365 licensing.

Risk signals appear in the registry's Risks column, aggregated from Entra, Defender and Purview by severity — they can lag the security portals by up to an hour and need E7 or Agent 365. Behind them:

  • Microsoft Defender — Assets > AI agents shows risk level, risk indicators, recommendations and active alerts; the advanced hunting table AgentsInfo holds agent inventory.
  • Microsoft Entra — sign-in logs can be filtered by agent type (agent identity, agent user, blueprint) and Is Agent; ID Protection's Risky agents report lets you confirm compromise (raising risk to High, which triggers Conditional Access policies that block high agent risk), confirm safe, dismiss or disable.
  • Microsoft Purview — the agent's Security tab links to the Purview Activity explorer.

⚠️ Exam Trap: Agent 365 activity numbers won't match the Copilot Agents usage report in Reports > Usage. That report counts agent use in Copilot Chat and the Office apps; Agent 365 covers every surface, including Teams and SharePoint. A mismatch is expected, not a fault.

Reflection Question: A partner agent shows a spike in exceptions and an "at risk" flag. Which three portals would you check behind the registry signal, and what action in Entra would immediately stop it signing in if you confirmed compromise?

See how it connects
Alvin Varughese
Written byAlvin Varughese
Founder•26 professional certifications earned