OpenAI agreed to stagger the rollout of its newest flagship model after a request from the Trump administration, the company said. Access to GPT-5.6 Sol begins with a limited preview for “trusted partners” cleared by the U.S. government before broader availability. OpenAI described the move as engagement with government security processes, while noting it does not want this kind of government-access framework to become a permanent default. The model will arrive alongside other tiers, including Terra and Luna, with OpenAI saying the rollout supports both performance and coding improvements. The staggered release underscores how frontier AI distribution is increasingly tied to U.S. security review—raising compliance and procurement considerations for universities and research centers that rely on new model capabilities on predictable timelines.