Microsoft announced new capabilities in Project Perception and Microsoft 365 E7 that treat enterprise AI agents as possible insider threats.
Guardian Agents supervise other agents' actions continuously, and policy controls adjust what each agent is allowed to do.
The framing matters: agents are now managed like employees with access, not like software features.
Our take
Agents as insiders is the right mental model. Expect every security vendor to copy it.
From the issue
The US just banned DeepSeek from federal agencies, two days after its best model yet Monday, October 19, 2026