Trump Administration Tightens Measures Around Anthropic; OpenAI Secures Pentagon Contract

Staff Report: PNN

Significant policy changes have emerged within the U.S. federal administration regarding the artificial intelligence company Anthropic. President Donald Trump, in a post on his social media platform Truth Social, directed federal agencies to stop using all products from Anthropic. He approved a six-month transition period but made it clear that in the future, Anthropic would no longer be considered a federal contractor.

Although Trump’s post did not explicitly label Anthropic as a ‘supply chain risk,’ later Defense Secretary Pete Hegseth separately announced the measure. According to him, in the interest of national security, Anthropic would be treated as a supply chain risk, and any contractor or partner affiliated with the U.S. military would not be allowed to engage in commercial activities with Anthropic.

The controversy began due to a disagreement between the Pentagon and Anthropic. The company refused to allow its AI models to be used for mass internal surveillance or fully autonomous lethal weapons. While a section of the Defense Department viewed these restrictions as excessive, Anthropic’s CEO Dario Amodei publicly stated that the company would not compromise on these two issues. He also mentioned that if the Defense Department desired, they would assist in a smooth transition to another supplier so that ongoing military plans or missions would not be disrupted.

Meanwhile, OpenAI publicly expressed support for Anthropic’s position. OpenAI CEO Sam Altman, in a message to employees, stated that they too are unwilling to engage in any illegal or inappropriate use—particularly involving internal surveillance and autonomous offensive weapons. This information was reported by BBC.

However, the situation quickly took a new turn. Within hours of the Trump administration ordering a severance of ties with Anthropic, OpenAI announced a new contract with the Pentagon. Altman claimed that even in this contract, the same policy boundaries regarding surveillance and autonomous weapons would remain in place. According to The New York Times, discussions between the government and OpenAI regarding potential partnership had already begun earlier in the week.

Meanwhile, OpenAI co-founder Ilya Sutskever commented on social media platform X that it is important to remain steadfast in Anthropic’s position, and OpenAI has adopted similar boundaries—this too is significant.

It should be noted that in July, Anthropic, OpenAI, and Google had each separately received contracts from the U.S. Department of Defense. Following the recent decision, new discussions have started concerning AI usage policies, ethics, and commercial competition in the U.S. defense sector. Observers suggest that in the coming days, more back-and-forth measures may be expected regarding this issue.