Trump’s Secret AI Rules Face Lawsuit for Secrecy

A nonpartisan group called Protect Democracy filed a lawsuit against four federal agencies, accusing them of keeping the rules for testing AI models completely secret. These rules decide which AI tools get approved for public use, but almost no details have been shared with Congress, the public, or even tech experts outside the government.

The lawsuit argues that this secrecy could lead to corruption, with officials potentially favoring companies they like or blocking others they dislike. For example, the Trump administration recently tried to blacklist an AI company for being ‘too woke,’ and a judge ruled that move illegal. Without clear rules, there’s no way to know if AI tools are being properly vetted for safety or security risks.

The lawsuit demands that the agencies release all the details of their review process by September 30, including who is involved, how they’re selected, and what criteria they use to approve or reject AI models. Protect Democracy warns that without oversight, the public could be left in the dark about whether these models are safe—or if they’re being used to favor certain companies.

Some lawmakers, like Rep. Greg Casar (D-TX), have criticized the Trump administration for what they call a ‘voluntary’ review process that no one can see. Meanwhile, California is considering its own AI safety standards that would be transparent and involve public input, unlike the secretive federal approach.