OpenAI said Friday that it is restricting the release of its newest artificial intelligence model at the request of President Donald Trump's administration, according to reporting carried by MSN and the Associated Press. The company has agreed to stagger the rollout so that the government can vet who gets access.
The model, identified by Shacknews as GPT-5.6, will be approved for customers on a case-by-case basis. As Fortune put it, OpenAI agreed to release its most powerful model only to Trump-approved customers, while the Financial Times reported the administration asked OpenAI to stagger the release specifically to vet users.
According to the AP item, the move is part of what it called an unprecedented government vetting of AI products that could pose cybersecurity risks. The Washington Post framed the upshot bluntly in its headline: the U.S. government will decide who gets to use the model. Politico reported the administration "stepped in" to limit the launch, and Barron's attributed OpenAI's narrowed rollout to federal involvement.
In plain terms, OpenAI is no longer the sole gatekeeper for its flagship product. Instead of releasing the model broadly the way past AI systems have shipped, the company will release it gradually, with the federal government screening prospective users during a cybersecurity review. Several outlets, including Startup Fortune, described this as a new government checkpoint sitting between the company and the public.
The sources do not specify how long the review will last, which agencies are conducting it, or the precise criteria for approval.
Why it matters: this is the first time the U.S. government has positioned itself as a direct gatekeeper deciding who can access a leading commercial AI model — a precedent that could reshape how powerful AI is distributed in the future.