Anthropic gives outside AI watchdog “employee-like” access

The outside watchdog can publish its findings without Anthropic's approval.

Source: Ars Technica · Summarized by HeadlinesBriefing