White House to Gatekeep Individual Access to OpenAI's GPT-5.6

The Trump administration has imposed a new policy on OpenAI requiring the company to stagger the release of GPT-5.6 and obtain White House approval for each individual customer. According to Axios and reporting by Stephanie Palazzolo, the government intervened because GPT-5.6 possesses "Mythos-like" capability—a reference to a previous frontier model with advanced security concerns. CEO Sam Altman told staff that the government will be approving access on a case-by-case basis, a highly unusual approach. Altman has made clear to the U.S. government that this is not OpenAI's preferred long-term model, but the company has no choice but to play ball.

The policy has drawn sharp criticism from analysts and industry observers. Andrew Curran notes that this does not slow development or training in any way—it only slows the rate at which models can be released to the public. The gap between what labs have internally and what is available to the public will steadily widen. This also means Chinese models are likely to face restrictions or even bans in the West, and if every American frontier release faces a slow, staggered rollout, Chinese alternatives—currently about nine months behind—could close that gap rapidly. Over time, this could destroy Western labs' business models as users turn to equivalent open-source alternatives. Critics describe the approach as "maximally not the way," arguing that ad hoc, opaque, politicized decisions from the White House on who gets frontier intelligence is a dangerous precedent for AI governance.