Melanie Mitchell asks how "independent" AI evaluators can escape the revolving door

MelMitchell1 · x · 2026-09-13

Replying to Nathan Lambert, Santa Fe Institute professor Melanie Mitchell raises a broader concern: how can "independent" AI evaluators avoid a revolving door with AI companies? If evaluators hope OpenAI might hire them or acquire their startups, how independent can they really be? She calls the institutional conflict-of-interest problem the real issue.

Related event: METR Investigator's Marriage to OpenAI Board Member Sparks AI Audit Independence Debate(6 posts)→

Original post →

More from Safety

Safety channel →