US grows anxious over AI, orders Sam Altman to delay GPT 5.6 rollout: Report
OpenAI has reportedly delayed the rollout of its upcoming GPT 5.6 AI model after the US government asked the company to make it available to a limited number of trusted partners first.
According to a Bloomberg report, OpenAI CEO Sam Altman told employees during an internal meeting that the Trump administration had requested a phased launch for the GPT 5.6 model.
OpenAI chief reportedly told employees that the US government is becoming more concerned about the capabilities of advanced AI models.
The latest news comes just weeks after the Trump administration ordered Anthropic to shut down access to premium Mythos 5 and Fable 5 models for foreign nationals inside and outside the US.
How will GPT 5.6 be released?
The report notes that OpenAI will initially gain access to approximately 20 trusted partners, with Amazon Bedrock expected to serve as one of the primary platforms for access to the model. OpenAI’s chief reportedly told employees that officials should work with the Trump administration on any input on the security and restrictions of OpenAI’s future models, even if the company disagrees.
Reportedly, OpenAI employees are also concerned about whether the government’s actions against Anthropic will affect the company’s ability to launch future AI models on time.
Anthropic has previously said it believes the government took action after researchers demonstrated a jailbreak that allowed Fable 5 to bypass some security guardrails for cybersecurity-related tasks.
“We disagree that a narrow potential jailbreak finding should be a reason to recall a commercial model deployed to hundreds of millions of people. If this standard were implemented industry-wide, we believe it would essentially halt all new model deployments for all leading model providers,” Anthropic said in a blog post.
The report notes that AI companies currently do not have a standard process for informing US authorities about the launch of their upcoming flagship models. The companies are also reportedly unclear about who in the government should be briefed, what technical information should be shared, and how far in advance officials should be notified before the launch.
However, the uncertainty could ease within a few months, with President Donald Trump signing an executive order earlier this month directing the administration and AI companies to develop a voluntary framework for frontier AI models within 60 days.
The proposed framework would allow the U.S. government to review certain advanced AI models for up to 30 days before their planned public release, creating a more structured approval process for future launches.
Trump administration will decide on partners
According to The Information, OpenAI will initially release GPT 5.6 as a limited preview for a small group of enterprise customers. It is reported that during this period, the Trump administration is expected to approve customer access on a case-by-case basis.
“We have made it clear to the US government that this is not our preferred long-term model and that we will work with them and others in the industry to achieve a more sustainable approach for future releases,” Altman said, as quoted by The Information.
