Anthropic's $7.7 Billion Stake Turns Its Safety Evaluator Into a Payroll Employee
Regulatory Capture by Anthropic Through METR
Kevin Bass claims an audit of Anthropic's finances reveals a self-reinforcing loop: METR, the third-party evaluator Anthropic proposes to assess its models, depends financially on Anthropic stock donated by Dustin Moskovitz — a stake that grew from roughly $500 million to over $7.7 billion in about 16 months. The same funders back groups that amplify AI-doom messaging, so Bass argues neither METR nor the doom machine can offer independent judgment, and calls for a Congressional investigation.
The "third-party evaluator" is not "third-party" at all. The evaluator is on Anthropic's payroll.
- m-hodges
Who is this person? His PFP is in a white coat and his bio says, “Kicked out of TTU med for truth-telling — I want Texas Tech to tell the truth about what it did”. This whole thing is clearly written by a language model.
- jrflo
None of this is new or enlightening information. Everyone has known that Anthropic and METR have a close financial relationship, that's been discussed at length if you follow AI development at all. Of course people investing in AI companies will invest in other AI companies, and many of the existential risk people at Anthropic will donate to AI safety causes...
- vb-8448
From an ethical standpoint is definitely something that need to be addressed.
Probably it's legal, but how can you trust their evaluation of the competition if that can hurt their wallet?
- aucls1
- flawn
AI-generated slop, this diagram was public for longer and has been posted here, iirc.