OpenAI has agreed to stagger the public release of GPT-5.6, limiting initial access to a small group of government-approved partners after a request from the Trump administration, according to a report from The Information published June 25. The arrangement is one of the clearest examples yet of federal officials shaping how and when a frontier AI model reaches the public.

How the rollout will work

CEO Sam Altman told employees during an internal Q&A on Wednesday that GPT-5.6 would launch first as a limited preview for a select group of enterprise partners rather than a broad public release. In a follow-up memo, Altman said the government would be "approving access customer by customer during this preview period," with a wider rollout expected a "couple of weeks" later, depending on how the approval process proceeds.

The request reportedly came from two federal bodies — the Office of the National Cyber Director and the Office of Science and Technology Policy — and was framed around the model's advanced capabilities and potential national-security implications, particularly in cybersecurity.

A new posture toward frontier releases

The move follows a broader policy shift. A June executive order asked frontier labs to give the government 30 days of early access to new models before release, while stopping short of mandatory licensing — the access is voluntary. The GPT-5.6 arrangement turns that early-access concept into something more hands-on: a government role in deciding which customers can use a model during its preview window.

That is a meaningful departure from the industry's usual pattern, in which labs control their own staged rollouts and decide who gets access. It raises practical questions that remain unresolved — what criteria the government uses to approve a customer, how long reviews take, and whether the same process will apply to rival labs releasing comparable models.

OpenAI's reservations

OpenAI complied, but Altman signaled the company sees the setup as temporary. According to The Information, he wrote that OpenAI has "made clear to the U.S. government that this is not our preferred long-term model, and will work with them and others in industry to achieve a more sustainable approach for future releases."

For the wider field, the episode is a marker worth tracking. If government screening of early access becomes a norm rather than a one-off, it would reshape release timelines and competitive dynamics across every frontier lab — not just OpenAI.