Anthropic resumes Pentagon talks as CEO Dario Amodei makes ‘last-ditch’ bid to avoid supply chain ban: Report
Anthropic CEO Dario Amodei is making a “last-ditch” attempt to negotiate a deal with the US Department of Defense after the company was blacklisted by the federal government last week, the Financial Times reported.
The report states that Amodei was in talks with Emil Michael, the undersecretary of defense for research and engineering, to finalize the contract for the Pentagon’s access to Anthropic’s artificial intelligence models.
It was notable that Michael called Amodei a “liar” with a “god complex” in a post on X last week.
“He wants to play God, make new ‘laws’ and… really stupid laws,” Michael said. “We never wanted the guardrails to be removed (a lie!) and accepted human control of all weapon systems.”
In particular, US Secretary of Defense Pete Hegseth had said that the US would identify Anthropic as a supply chain risk; this definition is often specific to foreign competing companies.
Anthropic said it would challenge this claim in court. The company said its talks with the Pentagon broke down after two conditions were set for Claude’s use in military use cases: no mass domestic surveillance and no use of fully autonomous weapons.
Why did Anthropic’s talks with the Pentagon fail?
In an internal memo to staff, Amodei reportedly explained why talks with the Pentagon had failed: “Towards the end of the negotiations, [department] We offered to accept our current terms if we deleted a specific phrase about ‘analysis of collectively obtained data’, which was the only line in the contract that exactly matched the scenario we were most concerned about. “We found this very suspicious.”
Amodei also noted that Anthropic stayed out of the controversy because “we don’t do dictator-style praise for Trump.” Anthropic’s CEO also claimed that many of the messages sent by the Pentagon and OpenAI “directly lie or attempt to confuse these issues.”
Anthropic became the first company to do so, signing a $200 million contract with the US government last year to deploy AI models in covert environments. But last week, President Donald Trump lashed out at the company for what he called “selfish” behavior and blacklisted the AI startup by the federal government.
Trump wrote: “The leftist lunatics at Anthropic made a HUGE MISTAKE trying to STRONGLY SHARE the War Department and force them to abide by the Terms of Service instead of our Constitution. Their selfishness puts AMERICAN LIVES at risk, endangers our Troops, and ENJOYS our National Security.”
While Trump added that the US government would not do business with Anthropic again, a report published a few days later claimed that the Pentagon used Anthropic’s artificial intelligence model in the recent attacks on Iran. It was also reported that the company’s artificial intelligence was last used during the capture of Venezuelan President Nicolas Maduro.
