Viral Report Questions Anthropic AI Safety Practices

A viral report is raising serious questions about the relationship between AI giant Anthropic and its primary safety evaluator, METR. Researcher Kevin Bass claims that Anthropic uses media narratives about potential AI risks to boost its valuation. He alleges this strategy benefits billionaire Dustin Moskovitz, whose equity in the company ultimately supports organizations that fund METR.

The report has drawn significant attention online, with some critics comparing the situation to the complex financial structures of failed exchange FTX. While there is no evidence of theft or illegal activity, the core concern centers on whether Anthropic and its safety overseer are too closely connected to provide truly neutral oversight. METR maintains that it strives for independent funding and does not accept direct payment for safety assessments from the labs it evaluates.

This controversy comes just as Anthropic CEO Dario Amodei is pushing for new safety measures, including the use of embedded evaluators within tech firms. Critics argue these moves are less about genuine safety and more about securing a form of regulatory capture that favors established players. As the conversation grows, some are calling for a formal Congressional investigation to clarify these financial links.

For the crypto and tech sectors, this situation highlights the rising scrutiny on the powerful entities shaping artificial intelligence. Investors should keep an eye on how these allegations impact public perception and future regulatory approaches to AI safety audits.