OpenAI’s New GPT-5.6 Sol Model Restricted to Trump-Approved Partners Only

ChatGPT developer OpenAI announced on Friday that, at the request of President Donald Trump’s administration, it will restrict the release of its latest AI model, GPT-5.6 Sol. The product is currently available only to “a small group of trusted partners approved by Trump,” marking the latest case of unprecedented U.S. government scrutiny of AI products that may pose cybersecurity risks.

OpenAI stated that this restriction is part of a temporary safety testing phase, designed to align with a government 30-day national security risk assessment process. An AI regulatory executive order signed by Trump in early June requires that the most advanced AI systems undergo federal review before public release. The company described this limitation as “a temporary step toward broader availability within weeks.” Previously, several tech executives had consulted with the White House on AI safety issues.

Industry observers believe this move signals a substantive intervention phase in U.S. oversight of frontier AI technology. Although OpenAI emphasizes that access will gradually expand after the testing period, the government’s direct designation of approved users has sparked debate over the balance between technological openness and security review. Over the next month, the test results of GPT-5.6 Sol and subsequent policy directions will become a key focus for the industry.