OpenAI's Joe Daroo: independent AI auditors need diverse expertise, not just AI safety

Dr_Atoosa · x · 2026-09-13

Joe Daroo of OpenAI publicly urged prospective independent evaluators auditing OpenAI, Anthropic and GDM to hire researchers across cybersecurity, bio, child safety, natsec and other domains—not just "AI safety"—so audits cover a wider sample of harms. He worries the field lacks diversity of expertise and welcomes progress on frontier safety collaboration.

Related event: Sriram Krishnan's Call to Fund a Distributed Ecosystem of Independent AI Evaluators Sparks Debate(15 posts)→

Original post →

More from Safety

Safety channel →