BREAKING: Secret US AI evaluation framework has been partly revealed

Substack · Watch/Read · (gathered 2026-09-16)

• The US government, under Trump, is secretly regulating “frontier” AI models through an undisclosed “AI security framework,” despite public claims against AI regulation.
• A FOIA request and lawsuit by Protect Democracy resulted in 132 heavily redacted pages, revealing almost no substantive details about the framework beyond the involvement of Michael Kratsios and Ethan Klein.
• The continued opacity of this AI evaluation policy raises concerns about potential “corruption and autocratic abuse,” with Protect Democracy vowing to continue its fight for transparency.