OpenAI is putting limits on its most capable artificial intelligence and spacing out how it ships new versions, following a request from the US government.
According to Computerworld, US authorities told OpenAI to restrict access to its most powerful AI model. The Guardian reports that the company is staggering its model release after a request from the Trump administration, rather than pushing the technology out all at once.
The practical effect is already showing up in OpenAI's product timeline. 41NBC News reports that OpenAI is delaying the release of ChatGPT 5.6 as AI oversight expands.
Taken together, the three reports describe a shift in how the most advanced AI reaches the public. Instead of a single, wide launch, the most powerful capabilities are being gated and rolled out gradually, with the government playing a direct role in shaping the pace.
The sources do not spell out the specific national-security or safety reasoning behind the request, the exact restrictions placed on the model, or a firm new date for ChatGPT 5.6. What is clear from the reporting is the direction: tighter access at the frontier, and a slower, more controlled release schedule.
Why it matters: how quickly the most powerful AI reaches everyday users — and who decides — is now being negotiated between a leading AI company and the US government, a sign that frontier AI is increasingly treated as something to be managed, not just launched.