Lack of Transparency in Government Frontier Model Safety Reviews
July 9, 2026
The specific dialogue and processes used between the government, Anthropic, and OpenAI to determine model safety remain undisclosed. The exact criteria for certifying frontier models as safe for public release are currently unclear.
HOW THIS AFFECTS YOU
●
policyYou should advocate for more transparent safety evaluation frameworks between regulators and developers.