The U.S. government requires OpenAI's GPT-5.6 to be released in a limited preview, with customer-by-customer approval for access

After a meeting between OpenAI CEO Sam Altman and U.S. Commerce Secretary Howard Lutnick, the U.S. government asked OpenAI to restrict the initial release scope of the GPT-5.6 model, making it available only to select partners, with access requiring individual approval on a per-customer basis, due to the model's advanced capabilities.
This intervention also involves the Office of the National Cyber Director and the White House Office of Science and Technology Policy, among other agencies, which will participate in additional cybersecurity testing.
This approach is similar to previous restrictions imposed on Anthropic's models, reflecting the U.S. government's increasing regulatory oversight over the release of frontier AI models.