Four federal agencies now face a lawsuit over a secret framework the Trump administration reportedly uses to review frontier AI models before they’re released. The nonprofit Protect Democracy filed the suit Wednesday, arguing that almost nothing about the process has been shared with Congress or the public — not who’s involved, what criteria they use, or what legal authority backs the reviews.
The complaint targets what it calls the “unclassified procedural and contractual architecture” of the framework. Protect Democracy wants the text of the framework, participation terms, participant identities, and the criteria for granting or withholding access to cutting-edge models. They’re asking a court to force production of those records by September 30, plus an injunction preventing officials from improperly withholding anything that isn’t classified.
One detail stands out: Protect Democracy claims OpenAI has “negotiated a private agreement with the Federal Government to limit distribution of its cutting-edge AI models to government-vetted partners.” That’s a striking arrangement if true — a private company and the executive branch deciding together who gets access to frontier technology, with no public record of the terms. This echoes earlier concerns about regulatory capture in AI governance, where the companies being regulated help write the rules.
“The executive branch is now choosing which companies can release their products and which customers get access to a technology that could shape the future of not only American industry and national security, but economies and governments abroad as well,” the group said in its release. The lawsuit frames this as perhaps the defining policy question of the current White House — and one that’s been operating entirely in the dark. Whether the courts agree that transparency is legally required remains an open question, but the September deadline gives the agencies little time to mount a defense.
