Anthropic grants outside AI watchdog employee-like access to its systems

Anthropic has given an external AI watchdog access to its operations that is comparable to what its own employees receive. The arrangement lets the outside group examine the company's AI work from an independent vantage point. Notably, the watchdog is permitted to publish its findings without needing Anthropic's approval. The report describes the move as a step toward greater outside accountability for the AI developer.