OpenAI is limiting the release of its powerful new GPT-5.6 AI models at the request of the Trump administration, the company said Friday. The voluntary compliance, rather than a legally binding order, highlights the administration's increasingly direct role in managing advanced AI tools and raises questions about how the government balances national security with innovation.
After OpenAI previewed its GPT-5.6 models - Sol, Terra and Luna - for federal agencies, the government asked the company not to publicly release the tools. OpenAI said it was temporarily complying and working with agencies on a formal model-review process that President Donald Trump called for in his recent AI security executive order. The administration's request highlights a new dynamic in AI for Government, where voluntary cooperation is becoming as important as formal regulation.
"We don't believe this kind of government access process should become the long-term default," OpenAI said in a blog post. "It keeps the best tools from users, developers, enterprises, cyber defenders, and global partners who need them." The company added that it plans to release the models broadly "in the coming weeks."
Anthropic's precedent and the administration's shift
The administration's request to OpenAI follows a more aggressive move against Anthropic. The Commerce Department ordered Anthropic not to share its powerful models Mythos and Fable with foreign nationals, prompting Anthropic to cut off almost all of its customers' access to those tools. That export-control restriction marked a sharp reversal inside the Trump administration, which had spent 2025 arguing against AI regulation to avoid giving China an advantage.
Mythos's debut in April changed the White House's calculus, kicking off a significantly more restrictive approach. On the same day OpenAI revealed its voluntary cooperation, Anthropic announced that the government had allowed it to resume offering Mythos to "certain trusted partners."
OpenAI CEO Sam Altman told employees on Thursday that the Trump administration would be "approving access customer by customer during this preview period," The Information reported. OpenAI did not respond to a request for comment about why it complied with the government's request.
Why this matters for government professionals
For policy makers and government officials, the episode illustrates the evolving relationship between federal agencies and AI developers. Voluntary requests, rather than binding regulations, are becoming a tool for managing national security risks associated with advanced AI. Understanding the nuances of these interactions is critical. Resources like the AI Learning Path for Policy Makers can provide a foundation in the governance challenges that are now emerging.
Your membership also unlocks:
