Microsoft bolsters security to better control AI insider risks

Microsoft expands Project Perception to secure enterprise AI agents from “agentic insiders,” with real-time risk assessment and permissions control.