US Treasury bans Anthropic as OpenAI clinches Pentagon deal

The US Treasury said it is ending use of all Anthropic products, including the Claude platform, after President Trump posted on Truth Social directing federal agencies to immediately cease using the AI firm’s technology.

Anthropic, a US AI start-up, vowed to sue over what it called intimidation and said its technology should not be used for the mass surveillance of US citizens or in fully autonomous weapons systems.

The Pentagon had demanded that Anthropic agree to unconditional military use of its Claude models, and said it operated within the law, adding that contracted suppliers could not set terms on how their products were employed.

Hours later, Sam Altman, OpenAI CEO, announced a deal with the Pentagon to use its models under technical safeguards the Department of Defense had agreed to, with similar red lines to Anthropic.

🔗 Source: South China Morning Post

🧠 Food for thought

Implications, context, and why it matters.

A contract dispute escalated into an unprecedented government crackdown

The government went further than ending the Treasury’s use of Anthropic products. The Pentagon moved to tag Anthropic as a “supply chain risk,” a label usually aimed at foreign adversaries 1.
That label raises the stakes for defense contractors that rely on Anthropic. They may drop the firm or risk losing their own government work, a shift described as potentially “existential” for the company 2.
Before the ban, officials also threatened to use the Defense Production Act. The law can force companies to put government needs first, including providing technology without restrictions 3.
Anthropic said the step went beyond a routine contract end. It called the move intimidation and said it was unprecedented for a US company in talks with its government 4.

The dispute signals a potentially more coercive era for government and big tech

The clash may chill the US AI industry. A former Trump administration AI adviser said he “could not possibly recommend investing in American AI” under these conditions 1.
Applying a “supply chain risk” tag to a domestic firm could become a precedent. That opens the door for similar pressure on other tech companies 2.
OpenAI later reached an agreement after getting similar ethical red lines written into its contract. Those limits sat at the center of Anthropic’s dispute with the Pentagon 5.
The end result suggests the hardline approach focused on making an example out of one company that would not agree without conditions 2.