US Government Now Controls Who Gets Access to GPT 5.6
OpenAI is rolling out its latest model, GPT 5.6, under an unusual restriction: the US government must approve access on a customer by customer basis before anyone gets in. CEO Sam Altman confirmed this during an internal Q&A session on Wednesday, according to a report by The Information.
The GPT 5.6 government approval process grew out of talks with two federal agencies, the Office of the National Cyber Director and the Office of Science and Technology Policy. Altman said he expects a broader release “a couple of weeks later,” assuming the preview phase goes smoothly.
The intervention did not stop there. Even after OpenAI had already shared its limited release plans with senior officials, Altman received a call from Commerce Secretary Howard Lutnick, who warned the company not to move forward without sign-off from additional agencies.
Altman was blunt about where the company stands: “We’ve made clear to the US 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.” The pressure appears linked to the Trump administration’s executive order calling for voluntary government safety reviews of new AI models, with a particular focus on cybersecurity. The OpenAI situation makes clear how voluntary that process actually is.
The backdrop to all this is what happened with Anthropic. When Anthropic launched its Claude Fable model, the US government stepped in and forced it offline despite the company having worked with authorities beforehand to identify security risks. The fallout stemmed in part from Anthropic’s strained relationship with the Pentagon, after the company declined to provide access for domestic mass surveillance and autonomous warfare operations, leading to it being classified as a supply chain risk.
OpenAI is clearly watching that situation closely, and the GPT 5.6 government approval requirement signals that AI labs may now need to navigate a de facto licensing regime whether they want to or not. Talks about re-releasing Fable are still ongoing. A White House spokesperson told The Information, “The White House continues to collaborate with frontier AI labs to develop shared approaches for addressing the challenges of scaling this technology.”
