OpenAI limits new AI models to trusted partners request US government
OpenAI CEO Sam Altman speaks to reporters after meeting with U.S. House Minority Leader Hakeem Jeffries on Capitol Hill in Washington, DC, on June 3, 2026.
Brendan Smialowski | AFP | Getty Images
OpenAI announced three new AI models on Friday, saying it complied with a U.S. government request to initially limit its rollout to “a small group of trusted partners.”
The company said in a blog post that it “believes in broad access” and is working to make GPT-5.6 Sol, Terra and Luna models generally available in the coming weeks. OpenAI said it had previewed the models’ capabilities and shared its plans with the government ahead of Friday’s launch.
“We do not believe this type of government access process should become the long-term default,” OpenAI said. “It hides the best tools from the users, developers, organizations, cyber defenders, and global partners who need them.”
OpenAI did not disclose the names of partners who could use its new models.
The announcement comes two weeks after rival Anthropic announced it had to disable access to two of its latest models to comply with a law. export control directive From the Trump administration. Anthropic is in active talks with officials in Washington, D.C., but has not said when it expects the models to come back online.
The Trump Administration has taken a noticeably more hands-on approach to AI regulations since President Donald Trump signed an AI executive order earlier this month. The order, thin on specific details, asked AI developers to voluntarily allow the government to evaluate model capabilities before full release.
OpenAI said it was working with the Trump administration to help create a framework for such evaluations and develop “a repeatable process for future model releases.”
“We are taking this short-term step because we believe this is the strongest path to broader availability in the coming weeks,” OpenAI said. he said.
GPT-5.6 Sol, Terra, and Luna are named according to their ability tier. OpenAI said it was Sol’s strongest offering yet.
The model shows improvements across coding and biology, and OpenAI said it is the company’s most cybersecurity-capable model. The company said it was better to help users patch vulnerabilities rather than carry out peer-to-peer attacks, and that OpenAI still did not pass the “critical” cybersecurity risk threshold, defined as introducing “unprecedented new paths to serious harm.”
WRISTWATCH: OpenAI to limit GPT 5.6 rollout after call from Trump administration
