Regulatory  ·  2026-06-27

White House / ONCD + OSTP: Trump Administration Restricts GPT-5.6 Release to Government-Approved Partners

RegulatoryHigh impactUnited States
On 25–26 June 2026, the White House Office of the National Cyber Director (ONCD) and Office of Science and Technology Policy (OSTP) requested that OpenAI limit the rollout of its new GPT-5.6 model family (Sol, Terra, Luna) to roughly 20 government-vetted 'trusted partners' via the API and Codex — blocking general public access pending a cybersecurity and national security review. OpenAI confirmed on 26 June it agreed to the restriction, calling it a 'short-term step' while working with the administration on a cyber Executive Order framework and a 'repeatable process for future model releases.' OpenAI stated: 'We don't believe this kind of government access process should become the long-term default.' The restriction applies to all three GPT-5.6 tiers simultaneously, a departure from normal commercial release practice. The government review process stems from the June 2 Executive Order (EO 14409) 'Promoting Advanced Artificial Intelligence Innovation and Security,' which created a 30-day voluntary pre-release vetting window for 'covered frontier models.'
This is the second consecutive frontier model gated by US government action (following the June 12 EAR export-control directive on Anthropic Fable 5 and Mythos 5) and marks a precedent-setting pattern: the US government is now actively vetting — and restricting — the commercial release of frontier AI models on national security and cybersecurity grounds before the public can access them. Enterprises, developers, and overseas partners that expected access to GPT-5.6 capabilities are blocked until the review concludes. The emerging framework treats the most capable AI models as quasi-controlled items requiring government clearance, reshaping how AI labs must plan product launches.
AI developers planning product launches on GPT-5.6 must apply through the government-vetted 'trusted partner' process or await broader release (expected within weeks per OpenAI). Legal and compliance teams should monitor the forthcoming Cyber Executive Order framework and the covered-frontier-model designation process under EO 14409 (30-day review clock ran to ~2 July 2026).
Sources
TechCrunch — OpenAI limits GPT-5.6 rollout after government request (26 Jun 2026)Axios — Trump administration asks OpenAI to limit next model release (25 Jun 2026)Axios — OpenAI releases powerful new GPT-5.6 model under restrictions (26 Jun 2026)AP News — OpenAI limits ChatGPT product to Trump-approved customers (26 Jun 2026)VentureBeat — OpenAI unveils GPT-5.6 Sol, Terra and Luna (26 Jun 2026)OpenAI X/Twitter confirmation — limited preview at US government request (26 Jun 2026)Politico — Trump administration steps in to limit OpenAI's latest model launch (26 Jun 2026)
See this in the live feed Explore related AI security and governance findings — updated every morning.
Open the feed →