OpenAI has rolled out its latest model, GPT-5.6, but it is not letting just anyone use it. According to Barron's, the company is restricting the limited release of the new model to the United States only.
Reporting from the Pittsburgh Post-Gazette and Audacy goes further, saying OpenAI is limiting its newest ChatGPT product to what they describe as "Trump-approved customers" while a cybersecurity review is underway.
The restriction does not appear to be happening in isolation. According to WION, the move follows a White House order issued two weeks earlier that directed another AI company, Anthropic, to block foreign nationals from accessing its Fable 5 and Mythos 5 models over national security concerns. In that framing, OpenAI is the second major US lab to have its most powerful model fenced off to domestic users.
On the product itself, SiliconANGLE reports that GPT-5.6 is positioned to challenge Anthropic's Claude Mythos 5. Coverage from MSN points to features reportedly named Sol, Terra and Luna, along with comparisons to the prior GPT-5.5. That same report cautions that OpenAI has not officially confirmed all the details, so some specifics remain unverified.
The sources here are largely aggregated headlines, and several leave open questions about exactly how access is being approved and what the cybersecurity review involves.
Why it matters: access to the most advanced AI is increasingly being treated as a matter of national security, and the US appears to be drawing borders around who can use frontier models — a shift that could reshape how AI is sold, shared and governed worldwide.