OpenAI has begun releasing its newest AI model, GPT 5.6, but is restricting initial access to a handful of approved partners, following guidance from the Trump administration. CEO Sam Altman reportedly informed employees that the government will be vetting each recipient individually, a process he described as authorizing one user at a time. This limited rollout comes after President Trump signed an executive order earlier this month encouraging certain artificial intelligence developers to voluntarily submit their models for federal security review before public release. The administration is currently evaluating GPT 5.6 over safety concerns, with multiple government offices requesting direct access to the system. OpenAI expects this preview phase to last roughly two weeks before potentially opening the model to wider use.