The Trump administration has asked OpenAI to stagger the release of GPT-5.6 over concerns about advanced AI models, according to a person familiar with the matter.

Sam Altman said OpenAI would first offer the model to 20 trusted partners through Amazon Bedrock. He also told employees to work with government input on safety and restrictions even when the company disagrees.

A White House official said the administration is working with frontier AI labs on shared approaches to AI risks, a stance that fits an existing federal framework under which advanced AI can require coordinated action..

The request came after Anthropic suspended global access to Mythos 5 and Fable 5 this month after a US order restricted foreign nationals inside and outside the US from using the models.

🔗 Source: Bloomberg

Recent OpenAI developments