What is Microsoft Guardian Agents (Defender agent discovery, dynamic permissions)?

Companies are giving AI agents access to their files, tools and systems. Microsoft argues that an agent acting with its user's access can behave like an insider threat, a risk that comes from people who already have legitimate access, and that with thousands of agents, manually triaging every alert will not scale. Guardian Agents are AI agents that Microsoft describes as helping security teams supervise other AI agents, investigate why risk changed, and recommend or apply containment. Dynamic Permissions are a companion control: Microsoft says an agent should begin with no authority and receive only what an approved task needs, for only as long as the task requires. Microsoft announced both on 19 October 2026 as extensions of Project Perception, its system of red, blue and green security agents.

Written from primary sources and reference pages, and checked against them by an independent AI fact-checker. Drafted with AI assistance.