Microsoft outlines security approach to the 'agentic insider' threat
In a blog post published on October 19, 2026, Microsoft focused on the 'agentic insider' risk posed by AI agents operating inside organizations. The company stresses that agents' permissions and access must be tightly governed.
