Anthropic CEO Dario Amodei appears to be caught between a rock and a hard place. Recently, he was forced to make a difficult decision over renewing the terms of a lucrative defense contract with the Pentagon. Now, it appears that Amodei may have to let go of the $200 million contract with the Department of Defense after his company was deemed a “supply chain risk” by the US security apparatus.
With Amodei refusing to lift the safety guardrails on his company’s AI model, Donald Trump has directed all federal agencies to adhere to a “six-month phase-out period” for the company’s products.
On the other hand, Amodei has made it clear that he is willing to cooperate with the US Department of Defense, provided it reconsiders its position on the safety clauses. To offer greater clarity on the situation, Jeffrey Ladish, an artificial intelligence and cybersecurity researcher and CEO of Palisade Research, weighed in, stating that both sides had reasonable demands.
Cybersecurity researcher explains Pentagon – Anthropic dispute
Ladish stated that Amodei genuinely wanted to help the US Department of Defense improve its security measures but was concerned that the company’s product could be “misused.”
“I think Dario [Amodei] is in a very difficult position… I know that he really does want to make sure that Anthropic’s AI systems are used to help improve U.S. national defense, but he’s also very concerned that those systems could be really misused,” the cybersecurity researcher said.
At the same time, he noted that the US government was within its remit to criticize Anthropic for not allowing unrestricted access to its AI model for “any lawful use” in matters of national security.
“I think it is reasonable for the government to be concerned about companies getting to dictate the terms of the most powerful technology humans have ever invented,” Ladish added.
Ladish further stated that it was not an overreaction on Amodei’s part to “thread the needle” with the US security apparatus, particularly given that China has deployed a massive surveillance system at scale that heavily relies on AI.
Notably, Amodei reportedly demanded safeguards against the use of its flagship AI model, Claude, for mass domestic surveillance and fully autonomous weapons systems. According to Amodei, he refused the contract because the legal language left room for the safety guardrails to be disregarded.
Amodei’s decision reportedly drew sharp criticism from officials within the Department of Defense, as well as from Donald Trump, who accused the company’s leadership of leaning too far into leftist ideology and attempting to strong-arm the department.
According to reports, a deadline — Friday at 5:01 p.m. Eastern Time — was provided by the Pentagon for Amodei to make a final decision. Ultimately, the two sides failed to reach an agreement, prompting US Secretary of Defense Pete Hegseth to state that “no contractor, supplier, or partner that does business with the United States military may conduct any commercial activity with Anthropic,” while designating the AI company a “security risk.”
Anthropic was reportedly one of several AI companies, including Google and OpenAI, to receive contracts worth up to $200 million from the Department of Defense last year. Its AI model was said to be the only one given clearance thus far for use in classified military missions. However, following the recent “phase-out” directive from Donald Trump, it remains unclear which other AI company will assume responsibility for providing models for national security purposes.
Although OpenAI CEO Sam Altman has stated that he has agreed to terms on a defense contract with the US government, there has been no official confirmation from the government at the time of writing. Reportedly, Elon Musk’s xAI has also been approved for use in classified military systems.
Read More: Donald Trump’s Top Defense R&D Official Calls Anthropic CEO A “Liar” With “God-Complex”
