OpenAI proactively restricts release of new AI models at government's request
OpenAI announced on Friday that, at the request of the Trump administration, it will restrict the public release of its new GPT-5.6 series models (Sol, Terra, Luna), offering limited previews to a small number of trusted partners first. The company stated it will temporarily comply with the government's request and plans a broader release within weeks. Previously, the government had imposed export controls on Anthropic's models, and this incident highlights a shift in AI regulatory policy.
OpenAI is restricting the release of its powerful new GPT-5.6 series AI models at the request of the Trump administration.
"We are starting with a limited preview for a small group of trusted partners, whose participation has been shared with the government, before a broader release," OpenAI said in a blog post on Friday.
OpenAI said that after previewing its GPT-5.6 models—Sol, Terra, and Luna—to federal agencies, the government asked the company not to publicly release these tools.
"We do not believe this government access process should become the long-term default," the company said in the blog post. "It keeps the best tools out of reach of the users, developers, businesses, cyber defenders, and global partners who need them."
OpenAI promised to publicly release the new models "in the coming weeks" and said it is temporarily complying with the government's request "because we believe it is the strongest path to broader availability in the coming weeks."
The company added that it is working with government agencies to advance the model review process required by President Donald Trump's recent executive order on AI safety.
The Trump administration's request to OpenAI comes after the government ordered Anthropic not to share its powerful models Mythos and Fable with foreign nationals, leading Anthropic to cut off almost all customer access to these tools.
The Commerce Department's export control restrictions on Mythos and Fable mark a dramatic shift in how the Trump administration is handling AI. Trump officials and advisers argued throughout 2025 that the government should not regulate AI, saying doing so would give China an advantage in the race to develop cutting-edge technology. But Mythos's debut in April changed the White House's calculus and ushered in a distinctly more restrictive approach.
OpenAI's situation is notable because the company is complying with a government request, not a legally binding order. According to The Information, OpenAI CEO Sam Altman told employees on Thursday that the Trump administration would "approve access on a customer-by-customer basis during this preview period."
OpenAI did not respond to a request for comment on why it is complying with the government's request.
On the same day OpenAI disclosed its voluntary cooperation, Anthropic announced that the government had allowed it to resume providing Mythos to "certain trusted partners."