OpenAI will reportedly delay the broad release of GPT-5.6 after a request from the Trump administration over potential security issues. The Verge, citing The Information, says the company plans to start with a limited preview rather than a wide rollout.
The episode shows how frontier model releases are becoming geopolitical and security events, not just product launches. Governments increasingly want visibility into risks before powerful new systems reach general availability.
For AI companies, staged deployment may become the default when models raise national security or misuse concerns.