Microsoft expanded Project Perception on October 19 to address risks from enterprise AI agents it describes as "agentic insiders." Microsoft's official blog argues that agents can hold broad access to ...