The White House is asking OpenAI to slow roll the release of its new model over safety concerns
TL;DR
OpenAI is reportedly not releasing GPT 5.6 broadly at first. Instead, the model would go only to a small group of close partners after pressure from the Trump administration. Sam Altman reportedly told staff that the government would approve access customer by customer during a preview period. A wider release could follow a couple of weeks later if the rollout goes well.
Nauti's Take
The issue is not that powerful models get reviewed. The issue is who decides access, and by which criteria.
If OpenAI and the government frame this as safety without giving meaningful detail, the result is a black box with major market impact. Safety should not become a polished label for opaque access, partner privilege, and political pre-selection.
Briefingshow
This is more than a routine safety review: it shifts part of frontier model control from companies to government agencies. For developers and customers, access to the most capable models could become more political, selective, and unpredictable. At the same time, it is still unclear whether the threat is proven or partly inflated for positioning.
Sources