US government reportedly asks OpenAI to slow GPT-5.6 rollout over safety concerns

The Trump administration has reportedly asked OpenAI to roll out GPT-5.6 in phases, starting with trusted partners. The move reflects growing US concern over advanced AI capabilities after restrictions imposed on Anthropic.

In Short

• US reportedly wants OpenAI to roll out GPT-5.6 in phases
• Government concerns over advanced AI capabilities continue to grow
• Anthropic's restrictions may have influenced the reported request

Just as the race to build more powerful AI models continues to accelerate, OpenAI has reportedly been asked by the US government to slow down the release of its next major model. According to a Bloomberg report, the Trump administration has asked OpenAI to stagger the rollout of its upcoming GPT-5.6 model instead of making it widely available immediately. The request comes nearly two weeks after rival AI company Anthropic suspended access to its most advanced AI models following regulatory pressure from the US government.

OpenAI asked to release GPT-5.6 in phases

According to the report, OpenAI CEO Sam Altman told employees that the US government wants the company to first release GPT-5.6 to a small group of trusted partners before expanding access to the public. During a meeting with employees, Altman reportedly said the government has become increasingly concerned about the capabilities of the most advanced AI models. He also told staff that OpenAI should work with the Trump administration on any feedback related to safety measures and restrictions for future AI models, even if the company disagrees with some of those suggestions.

The report says Altman did not specifically mention Anthropic during the meeting.

Anthropic's restrictions may have influenced the move

The reported request comes shortly after Anthropic shut off global access to its two most advanced AI models, Mythos 5 and Fable 5, after the US government ordered the company to restrict access for foreign nationals both inside and outside the United States, citing national security concerns.

Anthropic has since been in discussions with the US government in an effort to restore access to those models.

A letter reveals growing security concerns

This comes a day after a newly revealed letter sent by Anthropic to US Senators Tim Scott and Elizabeth Warren ahead of a Senate hearing on artificial intelligence accused China of distillation. In the letter, dated June 10, Anthropic alleged that Chinese technology giant Alibaba Group carried out a large-scale effort to gain unauthorised access to its Claude AI model by using thousands of fraudulent accounts.

According to the company, the campaign ran between April 22 and June 5, 2026, involving nearly 25,000 fraudulent accounts that generated more than 28.8 million interactions with Claude.

OpenAI employees are watching closely

According to people familiar with the matter, OpenAI employees are also concerned about how the government's actions against Anthropic could affect the ChatGPT maker's ability to launch future AI models more broadly.

- Ends