Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124

Anthropic CEO Dario Amodei publicly rejected the Pentagon’s request on Thursday. The Department of Defense wants unrestricted military use of the company’s AI technology. The deadline, which is just a few hours away, could lead to the $380 billion startup being kicked out of the US military’s supply chain.
The standoff marks the first time a major AI company has publicly challenged a US government threat to take control of its technology.
in Blog post Posting on Anthropic’s website, Amodei described the Pentagon’s threats as “inherently contradictory,” noting that one classifies Anthropic as a security risk while the other treats Claude as an essential element of national security.
“However, these threats do not change our position: we cannot in good conscience satisfy their request,” Amodei wrote.
The dispute centered on two conditions placed by Anthropy on the military use of Claude. The company prohibits self-targeting of enemy combatants and prohibits mass surveillance of American citizens. The Pentagon finds these restrictions on lawful military operations unacceptable.
Anthropic said the Pentagon’s “final offer,” which it received Wednesday night, did not resolve its main concerns. An anthropic spokesperson said in a statement, As reported by The Hill newspaper: “The new language written as a compromise has been coupled with legal terminology that allows those safeguards to be ignored at will.”
Defense Department spokesman Sean Parnell issued a public alert Thursday. He gave Anthropic until 5:01 pm EST on Friday to grant unlimited access to Claude Jouve — or face termination of the partnership and being labeled a supply chain risk.
“We will not allow any company to dictate how operational decisions are made,” Parnell wrote To X.
On Tuesday, Amodei met directly with Defense Secretary Pete Hegseth, where Pentagon officials outlined three consequences for non-compliance. First, remove regulations from military systems. Second, a supply chain risk classification that prevents other defense contractors from using manmade products. Third, invoke the Defense Production Act of 1950 to legally force the company to hand over its technology.
Amodei argued in the blog post that the rejection is also based on technical reality. “Frontier AI systems are not reliable enough to operate fully autonomous weapons,” he said, adding that without proper oversight, such systems cannot be trusted “to exercise the critical judgment that our highly trained and professional forces demonstrate every day.”
Republican Senator Thom Tillis criticized the Pentagon’s handling of the conflict. “Why the hell are we having this discussion in public? This is not how to deal with a strategic seller,” said Tillis. For journalists.
For Anthropic, the immediate exposure is a $200 million military contract. But the supply chain risk classification has much wider implications. Force every defense contractor to verify that they do not use human-made products in their operations.
The competitive landscape is changing rapidly. Elon Musk’s company xAI has signed an agreement to use Grok in classified systems. According to Axiosaccepting the “for all legal purposes” standard for classified works. OpenAI and Google accelerate negotiations to enter the ad space. Anthropic, previously the only authorized AI company for classified materials, risks losing its first mover advantage entirely.
The Pentagon’s willingness to enforce the Defense Production Act against a tech company sets a precedent that goes beyond artificial intelligence. If a government can legally force an AI company to remove security restrictions for national security reasons, the same framework could, in theory, apply to force cryptocurrency companies to modify privacy features or weaken transaction safeguards.
This comparison also strengthens the case for the development of decentralized artificial intelligence. A central AI provider could be pressured — or legally forced — to remove barriers to government demand. This confirms the hypothesis that decentralized alternatives provide a more resistant infrastructure against state coercion.
The rapid growth of anthropic has already raised concerns for the cryptocurrency markets. Company valuation amount to $ 380 billion AI-driven disruption to traditional software revenues is putting pressure on private credit flows that are closely tied to Bitcoin.
Anthropic also has a historical connection to cryptocurrencies: the bankruptcy estate of FTX had a significant stake in the company, which it later sold to help finance the repayment of creditors.
The deadline expires on Friday, but the real question begins then: will the Pentagon enforce compliance, and what it means for any technology company that draws a line between government contracts and product integrity.