WASHINGTON—The White House finalized a voluntary framework for artificial intelligence companies to submit advanced models for pre-release government review, but did so behind closed doors. The Aug. 1 deadline for public deliverables, established by Executive Order 14409, passed without any public announcement from the administration.
President Trump signed Executive Order 14409 on June 2, directing federal agencies to establish the framework. The order created a structured process for AI developers to share their most advanced models with the government before public release.
The private finalization follows the last-minute cancellation of a public signing for an earlier version of the order, which had triggered internal conflict among administration officials over its scope and provisions.
David Sacks, the administration's AI and crypto czar, played a central role in pushing back against the initial proposals. He told President Trump the provisions were a concern for the industry.
Sacks argued the framework's original language risked harming U.S. innovation and competitiveness against China's advancing AI sector. His intervention exposed divisions within the administration over AI oversight.
The completed framework is voluntary—a concession to industry interests that allows AI companies to choose whether to participate in the government's review process for their most advanced models.
The missed Aug. 1 deadline meant no formal notices appeared in the Federal Register and no public statements or publications from agencies including the National Institute of Standards and Technology or the Cybersecurity and Infrastructure Security Agency. Both were expected to contribute to and publish details of the framework.
The voluntary structure reduces immediate regulatory burden on AI companies and addresses concerns Sacks raised, but the absence of public guidelines leaves the sector without clarity on what participation entails or what the government expects.
The public has no specific details on how the government plans to assess the safety and security of advanced AI models before they become widely available. The White House has not said when it plans to release the framework details or how it will engage with the broader AI community.
