Pentagon Demands Anthropic Lift AI Safeguards or Lose $200M Contract
Defense Secretary Pete Hegseth has issued a Friday deadline to Anthropic, the AI company behind the Claude model, to loosen safety restrictions on its technology or risk losing a $200 million Pentagon contract.
The escalating dispute, reported by CNN and other outlets, centers on the Pentagon’s desire for unfettered access to Anthropic’s AI capabilities for “all lawful use,” according to two sources with knowledge of the discussions. Hegseth has also threatened to place Anthropic on a government blacklist, effectively limiting its ability to work with other defense contractors.
Anthropic, however, is reportedly unwilling to compromise on two key concerns: the development of AI-controlled weapons and the potential for mass domestic surveillance of American citizens. The company believes the technology is not yet reliable enough for autonomous weapons systems and that existing legal frameworks are insufficient to govern the use of AI in widespread surveillance, one source familiar with the matter said.
The Pentagon plans to terminate the contract if Anthropic doesn’t comply by 5:01 p.m. On Friday, according to a Pentagon official. Should Anthropic refuse, Hegseth intends to invoke the Defense Production Act, compelling the company to work with the Pentagon regardless of its objections. He also plans to designate Anthropic as a supply chain risk, a move that would effectively prohibit companies with military contracts from utilizing Anthropic’s products.
The Defense Production Act, originally used during the Trump administration to address pandemic-related supply shortages, grants the government broad authority to influence businesses in the interest of national defense. A supply chain risk designation is typically reserved for entities considered extensions of foreign adversaries like Russia or China.
Katie Sweeten, a former liaison for the Justice Department to the Department of Defense and now a partner at the law firm Scale, questioned the logic of simultaneously labeling Anthropic a supply chain risk while attempting to compel its cooperation. “I would assume we don’t want to utilize the technology that is the supply chain risk, right? So I don’t know how you square that,” Sweeten told CNN. “What it sounds like is that the supply chain risk may not be a legitimate claim, but more punitive because they’re not acquiescing.”
Despite the high stakes, a source familiar with Tuesday’s meeting between Hegseth and Anthropic CEO Dario Amodei described the conversation as cordial and respectful. Hegseth reportedly praised Anthropic’s products and expressed a desire to continue working with the company. However, Amodei reiterated Anthropic’s firm stance against loosening restrictions on autonomous weapons and mass surveillance, and the company has shown no indication of backing down.
A Pentagon official insisted the issue does not involve the actual deployment of autonomous weapons or mass surveillance, claiming the Pentagon “has always followed the law.” The official stated, “You can’t lead tactical ops by exception. Legality is the Pentagon’s responsibility as the end user.”
Anthropic, in a statement to CNN, characterized the meeting as a “good-faith” conversation. “Anthropic CEO Dario Amodei met with Secretary Hegseth at the Pentagon this morning. During the conversation, Dario expressed appreciation for the Department’s work and thanked the Secretary for his service,” the statement read. “We continued good-faith conversations about our usage policy to ensure Anthropic can continue to support the government’s national security mission in line with what our models can reliably and responsibly do.”
The negotiations have been ongoing for several months, with reports of tensions between the Pentagon and Anthropic surfacing in recent weeks. Last week, Axios reported that Hegseth was considering terminating the contract and designating Anthropic a “supply chain risk.”
Anthropic maintains its commitment to supporting U.S. National security through AI, noting it was the first frontier AI company to place its models on classified networks and provide customized models for national security customers. The standoff could potentially benefit Anthropic’s competitors, with a Pentagon official confirming that xAI, Elon Musk’s AI company, “is on board with being in a classified setting,” and others are “close” to reaching an agreement.
Anthropic has long positioned itself as a leader in AI safety, founded by former OpenAI employees who left over disagreements regarding the company’s approach to safety and the pace of AI development. The company recently announced a $20 million donation to a political group advocating for increased AI regulation.