Brundage: Mandate Audits at Leading AI Labs With Real Safety Standards

Miles_Brundage · x · 2026-09-16

Brundage lays out concrete demands: auditing should be mandatory at leading AI companies with minimum qualification, independence and access criteria, and regulators should enforce actual safety and security standards—not just "have a possibly weak policy and follow it."

Related event: Ex-OpenAI policy chief Brundage pushes for mandatory frontier AI audits(5 posts)→

Original post →

More from Safety

Safety channel →