New York Post questions close ties between METR and Anthropic

The New York Post reported on potential conflicts of interest between AI safety evaluator METR and Anthropic. METR, which advocates for deep access to AI companies for safety audits, shares historical and personal ties with Anthropic leadership, including shared past employment at OpenAI and personal relationships between staff and executives. METR stated it does not accept funding from the AI companies it evaluates, though it receives free model tokens.