OpenAI has released its newest family of AI models, GPT-5.6, moving them from a limited preview for select partners into public availability. According to Times Now, the lineup includes three variants named Sol, Terra, and Luna.
The rollout is drawing scrutiny in Washington. The Independent reported that OpenAI is launching GPT-5.6 amid US national security concerns. Both Digit and Times Now report that the White House has denied approving the public release, even as some reports claimed safety testing was completed before the models went live. In other words, the company shipped the models without the government sign-off that some accounts suggested had taken place.
On the business side, OpenAI says GPT-5.6 is now the "preferred model" for Microsoft 365 Copilot, according to TechCrunch. That means the new models will continue to power Microsoft's suite of workplace and productivity apps. TechCrunch notes this endorsement comes "amid breakup chatter" about the OpenAI–Microsoft partnership, making the continued integration a notable signal that the relationship remains intact for now.
On performance, the International Business Times reported that OpenAI CEO Sam Altman said the new series increases efficiency by more than 50%.
Why it matters: The launch highlights a growing tension between how fast AI companies release powerful new systems and how much oversight governments actually have — a gap the White House's denial puts in sharp relief.