OpenAI adjusted its release strategy for the GPT-5.6 series after discussions with the Trump administration, limiting initial access to government-approved partners rather than rolling it out broadly.
CEO Sam Altman acknowledged the shift in an internal memo, describing the controlled release as “not the preferred method.” But preferred or not, the company appears to have decided that working with Washington beats working against it.
What actually happened
On June 25, the Trump administration asked OpenAI to stagger the public release of GPT-5.6 over security concerns. The request came on the heels of an executive order issued earlier in June that established a voluntary 30-day review process for frontier AI models before they hit the broader market.
OpenAI complied. The company previewed GPT-5.6 for federal agencies first, with plans to expand global access by July 8 following government approval.
