OpenAI Limits Early Access to New GPT-5.6 Models at the Request of the U.S. Government

OpenAI is rolling out its new GPT-5.6 model lineup, but only a small group of trusted partners will get early access. The company made the move after the U.S. government requested that the launch begin with a limited preview.

The new lineup includes three models: the flagship Sol, the everyday assistant Terra, and the fast, more affordable Luna. OpenAI plans to make all three publicly available in the coming weeks, but for now, access is limited to a select group of trusted partners.

The company said it briefed the U.S. government on its plans and the capabilities of the new models before the launch.

“As part of our ongoing collaboration with the U.S. government, we shared our launch plans and model capabilities ahead of today’s release. At their request, we are starting with a limited preview for a small group of trusted partners whose participation was disclosed to the government before making the models more broadly available,” OpenAI said.

During the preview period, OpenAI will continue testing the models and coordinating with partners ahead of the wider public rollout. At the same time, the company stressed that it does not believe this type of government access process should become the long-term standard.

“We do not believe this government access process should become the long-term standard. It withholds the best tools from users, developers, businesses, cyber defenders, and global partners who need them,” OpenAI said.

OpenAI described the arrangement as a temporary measure and did not disclose which organizations were selected as trusted partners.

The move could mark the first public use of the advanced AI model review process introduced under a recent executive order signed by U.S. President Donald Trump. The voluntary framework allows developers to provide the U.S. government with secure access to advanced AI models up to 30 days before they are released to trusted partners.

OpenAI is not the only AI company to face increased scrutiny from the U.S. government. Earlier this month, Anthropic was forced to abruptly disable its Fable 5 and Mythos 5 models after the United States imposed export restrictions requiring the company to suspend access for all foreign organizations. It remains unclear when those models will become available again.

According to OpenAI, Sol is the company’s most powerful model to date, featuring enhanced agentic capabilities in coding, biology, and cybersecurity.