White House Requests Limited Rollout of OpenAI’s Next AI Model Over Safety Concerns
The White House has asked OpenAI to restrict the initial release of its upcoming GPT-5.6 artificial intelligence model to a small group of government-approved partners, according to a source familiar with the discussions.
The request follows growing concerns within the US government over increasingly powerful AI systems and comes shortly after the administration imposed export restrictions on Anthropic’s latest AI models, Mythos and Fable. Those models drew attention in Washington and on Wall Street because of their advanced cybersecurity capabilities, which some officials believe could pose significant safety and national security risks.
According to the source, both the US administration and OpenAI consider GPT-5.6 to possess capabilities comparable to Anthropic’s most advanced models. As a result, OpenAI agreed to limit the model’s initial deployment while broader regulatory discussions continue.
Technology publication The Information, which first reported the White House request, said OpenAI Chief Executive Sam Altman informed employees of the decision in an internal memo.
According to the report, Altman wrote that government officials are approving access to the model “customer by customer.”
He also acknowledged that the company does not view the arrangement as a permanent solution.
“We’ve 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.”
A White House official told CNN that the administration continues to work closely with leading AI developers to establish common approaches for addressing the challenges associated with increasingly powerful AI systems.
OpenAI later confirmed the limited rollout in a public statement, saying the decision was made at the government’s request.
The company, however, stressed that it does not believe restricting access through government approval should become standard practice.
“We don’t believe this kind of government access process should become the long-term default. It keeps the best tools from users, developers, enterprises, cyber defenders, and global partners who need them.”
OpenAI added that it hopes to make GPT-5.6 broadly available “in the coming weeks” while continuing discussions with government officials on a more predictable framework for future AI releases.
Regulatory Uncertainty Continues
The limited rollout comes despite an executive order signed earlier this month by President Donald Trump directing developers of advanced AI models to voluntarily submit their systems for government review 30 days before public release. However, a formal review process has yet to be established.
The absence of a clear regulatory framework has created uncertainty across the AI industry.
While the request to OpenAI originated from the White House, the recent export restrictions affecting Anthropic were issued by the US Department of Commerce, leaving companies unclear about which government body is ultimately responsible for overseeing advanced AI technologies.
Experts broadly agree that governments should play a role in addressing AI safety and national security concerns. However, many have warned that the current case-by-case approach lacks transparency and consistency.
Brad Carson, president of the bipartisan AI safety advocacy group Public First, argued that recent developments highlight the need for formal regulations.
“The Fable episode shows the need for clear regulations. Right now, you have an ad hoc, personalized, opaque, possibly lawless approach.”
He added that while governments have a legitimate role in restricting technologies deemed dangerous, such decisions should be made under a transparent and fair regulatory system.
“It is certainly appropriate for the government to recall dangerous products, including AI models, but it has to be done in a way consistent with transparency and basic fairness.”
