OpenAI rolled out a U.S.-only preview of its newest AI model series on Friday, giving a small group of trusted partners early access at the direct request of the U.S. government. The move comes just two weeks after the White House surprised Silicon Valley by ordering Anthropic to cut off all foreign nationals from its Fable 5 and Mythos 5 models over national security concerns, a directive Anthropic said it could not safely follow and responded to by shutting down access entirely.
Why Are AI Models Facing New Government Scrutiny?
The latest models from top AI firms, including Anthropic's Mythos series and OpenAI's new GPT-5.6, have raised concerns due to their reported ability to spot software vulnerabilities, the kind of coding weaknesses hackers often exploit.
Earlier this month, President Trump signed an executive order creating a voluntary federal review process for checking national security risks in advanced AI models before they hit the market. So far, the White House hasn't shared many details on how this review will actually work, including which models will be subject to it and how compliance will be enforced.
ALSO READ: WhatsApp's Encryption Bug That Revealed Blocks Has Finally Been Patched
The move is notable given the administration's broader stance favouring lighter AI regulation, even as it has worked to stop individual states from setting their own AI rules. Anthropic's case, in particular, has drawn criticism over alleged government overreach. OpenAI also said it wasn't entirely comfortable with the process this time around, but complied anyway.
"We don't believe this kind of government access process should become the long-term default," the company said in a blog post. "It keeps the best tools from users, developers, enterprises, cyber defenders, and global partners who need them."
What Models Did OpenAI Just Release?
The GPT-5.6 lineup includes three models: Sol, the new flagship; Terra, built for everyday business use; and Luna, a faster, budget-friendly option. OpenAI said Terra will eventually be priced at half the cost of its predecessor, GPT-5.5, as competition with Anthropic and Google heats up.
ALSO READ: iPhone 18 Pro, iPhone Ultra Launch Date Leaked With A Stinging Price Hike
Notably, Anthropic's earlier dispute with the Pentagon, after it refused to allow its tech to be used for surveillance or autonomous weapons, is still being fought out in court. Both OpenAI and Anthropic have filed confidential paperwork for IPOs, each eyeing valuations near $1 trillion.