When enterprises deploy Claude for Work, every agent inherits the employee’s full identity and access rights with no guardrails, no policy, and no audit trail. Here’s how one $300B investment firm closed that gap, and what it means for yours.
AI agents are no longer just chatbots. They’re executing multistep workflows across tools and data sources, and the Model Context Protocol (MCP) standardizes these interactions.
For years, artificial intelligence has been reactive. You prompted it, and it responded by analyzing data, generating text or predicting outcomes, but only when asked.
Anthropic’s disclosure of an AI-driven espionage campaign it halted is best understood as a faster, more persistent version of patterns the industry has seen before. What distinguishes this incident is the continuity of activity an autonomous system can sustain once it is given the ability to interpret its surroundings and act on that understanding.
Based on responses from more than 200 enterprises, the findings show how AI agents are reshaping identity attribution and access control in ways existing models were not designed to handle.
By combining identity-based access control with content inspection, this closes a gap most teams are still trying to manage with separate tools and after-the-fact controls.
