Trump may be forced to reveal secret rules feds use for AI safety testing
2026-09-02 · Ars Technica
Trump administration is keeping its AI safety testing rules classified, because nothing says 'transparent governance' like secret reviews of the most powerful AI models. A lawsuit is now trying to force disclosure, alleging the secrecy could be hiding corruption in how frontier AI gets evaluated. When your AI safety process itself needs a safety review, you know things are going great.