MP.AIGOV.1052 Agents should not have active risk indicators in Microsoft Defender
Microsoft Defender assesses each agent and flags real exposure, such as running on a vulnerable device or sitting unused while keeping its access. These indicators point to the agents most likely to be misused or attacked, so they should be dealt with first.
Remediation action
In the Microsoft Defender portal, go to Assets > AI agents, open each agent listed, and review its risk indicators, evidence and security recommendations. Fix the cause (patch the device, retire a dormant agent, reduce what a published agent can reach) or block the agent.
Related links
- Discover AI agents and assess security posture using Microsoft Defender
- AI agent posture risk in Microsoft Defender
- AgentsInfo table
- How this check works
%TestResult%