In a dramatic pivot, Anthropic has reportedly returned to the negotiating table with the U.S. Department of War in a last-minute effort to secure continued access to Pentagon contracts as the company faces the possibility of being labeled a supply chain risk by the Trump administration.
The move comes less than a week after President Trump ordered all federal agencies to “immediately cease” using Anthropic’s products, following a breakdown in talks over the military’s use of its Claude AI models.
Anthropic Reworks Pentagon Deals Over Trump’s “Foreign Adversary” Label
According to a report by the Financial Times, the company’s CEO, Dario Amodei, has been holding discussions with the U.S. Undersecretary of Defense for Research and Engineering, Emil Michael, to finalize terms governing the military’s use of its models. A new agreement would allow the Pentagon to continue using Anthropic’s technology, preventing a formal designation that would force contractors in the defense supply chain to cut ties with the AI company.
Sources familiar with the matter suggest the two sides are reviewing a compromise that would allow the Pentagon to utilize Claude’s advanced reasoning models for “mission-critical” intelligence – similar to its use in the January operation to capture former Venezuelan President Nicolas Maduro – while maintaining technical safeguards against domestic misuse.
The talks come after a breakdown in negotiations last week, when both sides failed to agree on language Anthropic said was necessary to prevent misuse of its technology. In an internal memo to staff, Amodei reportedly wrote that near the end of negotiations, the Pentagon offered to accept Anthropic’s broader terms if the company removed a clause restricting the “analysis of bulk acquired data.” He said this phrase was meant to guard against potential mass domestic surveillance, a scenario that the company treats as a red line. Additionally, the firm held firm against the deployment of its models in fully autonomous lethal weapon systems.
The crisis reached a fever pitch last Friday, when U.S. Secretary of Defense Pete Hegseth warned the San Francisco-based startup could be designated a supply chain risk to national security. That label, historically reserved for adversarial foreign entities, like Huawei and Kaspersky, would effectively freeze Anthropic from doing business with the U.S. military and its contractors – a move that could isolate it from partners like Amazon, Google, and Nvidia.
The standoff came despite Anthropic’s existing ties to the Department of War. The company was awarded a contract worth up to $200 million by the federal government in July 2025, and it became the first AI provider whose models were used in classified environments and by national security agencies. The U.S. military even used Claude to support a major air strike on Iran hours after Trump ordered federal agencies to stop using Anthropic’s systems.
Silicon Valley Warns Tech Crackdowns Hurt U.S. in Global AI Race
However, the Trump administration’s threats have sparked a rare wave of unity across Silicon Valley. In a Wednesday letter to the President, tech groups Software & Information Industry Association (SIIA), TechNet, Computer & Communications Industry Association (CCIA), and Business Software Alliance warned that labelling a domestic AI company a supply chain risk could undermine U.S. leadership in the field.
The signatories argued that treating a U.S. technology company as a “foreign adversary” rather than an asset could discourage innovation and weaken America’s ability to compete with China in the global AI race. These organizations represent hundreds of American tech firms, including Nvidia, Google, and Apple.
