**OpenAI Restricts ChatGPT Rollout at Request of US Government**
OpenAI has announced a restriction on the initial rollout of its latest AI models, a decision made at the request of the US government. This marks a significant change in the company's approach, as previous versions of ChatGPT were released directly to the public according to OpenAI's own timeline.
The recent decision follows a similar action taken by Anthropic, another AI research company, which disabled access to its newly launched models, Fable 5 and Mythos 5, just three days after their release. This move was also prompted by a directive from the US government concerning export controls tied to national security issues.
In a statement issued on Friday, OpenAI confirmed that it has initiated a "limited preview" of its GPT 5.6 series. This preview is being made available to a small group of trusted partners, with the identities of these participants shared with government officials. OpenAI described this arrangement as “short-term,” aimed at bridging the gap while the administration develops a framework for reviewing advanced AI models, as outlined in an executive order signed by President Donald Trump on June 2.
Despite the current limitations, OpenAI has expressed its commitment to broad access to its technology. The company plans to make its new models—Sol, Terra, and Luna—generally available in the upcoming weeks.
Reports from Axios and The Verge indicate that Anthropic has managed to restore access to its Mythos 5 model on a limited basis after addressing the government's concerns. However, restrictions on the Fable 5 model remain in effect. US Commerce Secretary Howard Lutnick noted in a letter dated June 26 to Anthropic co-founder Tom Brown that the company’s cooperation with authorities has led to "significant progress."
This tightening of oversight has raised concerns among some media outlets and policy commentators, who argue that the US government is extending its influence over the deployment of advanced AI systems without a clear legal framework. Critics warn that such ad hoc oversight could create unpredictability for developers and potentially undermine the competitiveness of US AI companies in the global market.
The recent shift in the administration's stance on artificial intelligence represents a notable change from earlier policies. After returning to office in 2025, President Trump had advocated for minimal regulation of AI technologies, referring to AI as "a beautiful newborn baby" essential for competing with China. He cautioned against "politics or stupid rules" that could impede the growth of the sector.
However, in light of rising cybersecurity and national security concerns, the administration has begun to adopt a more cautious approach towards frontier AI models. As these technologies continue to evolve and become more capable, the government appears to be prioritizing oversight to ensure that national interests are safeguarded.
The implications of this new regulatory environment for AI developers and companies remain to be seen, as the balance between fostering innovation and ensuring security becomes increasingly complex. OpenAI's decision to restrict its rollout may set a precedent for how future AI technologies are managed in the United States.