U.S. President Donald Trump said it is “possible” that Anthropic could reach an agreement with the Department of Defense to deploy its AI models, signaling a potential reversal after months of conflict. Speaking to CNBC, Trump pointed to recent discussions with the company, describing them as “very good talks” and suggesting progress toward a deal.
The remarks follow a high-profile dispute earlier this year, when the Pentagon labeled Anthropic a supply chain risk and restricted the use of its Claude systems in defense-related work. Trump also ordered federal agencies to stop using the company’s technology, escalating tensions between the government and the AI startup.
Despite those restrictions, the Department of Defense continued limited use of Anthropic’s models in certain contexts, highlighting the complexity of enforcing such bans during ongoing operations.
From Conflict to Negotiation
The relationship began to shift after Anthropic CEO Dario Amodei met with senior administration officials, including Susie Wiles and Scott Bessent. The White House described the discussions as “productive and constructive,” indicating a softer stance from policymakers.
A key factor in renewed interest is Anthropic’s recently unveiled Claude Mythos Preview, part of its Project Glasswing initiative. The model has drawn attention for its advanced cybersecurity capabilities, which could be valuable in defense applications. Anthropic has limited access to the system while continuing discussions with government agencies.
Earlier negotiations between the Pentagon and Anthropic collapsed over disagreements about how the technology could be used. The Defense Department sought broad access to the models, while Anthropic pushed for restrictions preventing use in autonomous weapons or domestic surveillance.
Strategic Stakes for AI and National Security
The potential deal highlights the growing importance of AI in national security and defense. Governments are increasingly seeking access to advanced AI systems for cybersecurity, intelligence, and operational support, while companies are navigating ethical and regulatory concerns.
Anthropic, founded in 2021, had previously secured a $200 million contract with the Pentagon before the dispute disrupted collaboration. The company later challenged the government’s restrictions in court, and parts of the administration’s directive have been temporarily blocked.
Trump’s comments suggest a possible path toward renewed cooperation, as both sides reassess the balance between technological advantage and risk. Any agreement would likely include safeguards governing how AI systems are deployed in military contexts.
The outcome of these negotiations could shape how AI companies engage with governments, particularly in sensitive areas such as defense, where the line between innovation and risk remains under close scrutiny.