White House Asks OpenAI to Limit GPT 5.6 Model Release to Government-Approved Partners Over National Security Concerns
The White House asked OpenAI to limit the GPT-5.6 rollout to a small set of government-approved partners, with Axios reporting the request came specifically from the Office of the National Cyber Director and the Office of Science and Technology Policy
The White House has installed itself as an informal pre-release clearance authority for frontier AI, operating through ONCD and OSTP with no statutory basis and no consistent process. OpenAI's public dissent, calling the arrangement not its "preferred long term model," signals compliance is strategic rather than durable. The "customer by customer" approval pathway cannot scale at commercial velocity, and with no enforcement mechanism the arrangement depends entirely on voluntary cooperation from a company under acute pressure from Chinese open-source rivals. Moderate confidence rests on four independently-placed primary sources converging on the same factual core. OpenAI may instead have sought the restriction proactively, using government imprimatur as competitive cover to delay rivals' enterprise access while building a compliant access tier.
6 sources
- Trump administration asks OpenAI to limit next model release over security concerns -
Axios - OpenAI limits new AI models to 'trusted partners' at request of U.S. government -
CNBC - White House asks OpenAI to limit its next model release -
CNN - OpenAI limits GPT-5.6 rollout after government request, says restrictions shouldn't be the norm -
TechCrunch - Trump Administration Asks OpenAI to Stagger Release of New Model Over Security Concerns -
The Information - White House asks OpenAI to limit its next model release -
CNN