President Donald Trump ordered federal agencies Friday to begin phasing out their leverage of technology developed by Anthropic, an artificial intelligence company, following a dispute with the Pentagon over access to its Claude AI model. The order, announced on Trump’s Truth Social account, mandates that most agencies cease using Anthropic’s products immediately, while granting the Department of Defense a six-month window to transition away from the technology.
The conflict stems from the Pentagon’s attempt to compel Anthropic to grant unrestricted military access to Claude, its flagship AI model. Defense officials sought full lawful access without the corporate-imposed usage restrictions Anthropic had implemented, according to a statement released by the Department of Defense on February 26, 2026. This “best and final offer,” delivered the evening prior, carried a deadline of Friday.
Anthropic resisted the Pentagon’s demands, prompting Defense Secretary Pete Hegseth to consider invoking the Defense Production Act, a Cold War-era law that allows the government to compel private companies to fulfill national security needs. Hegseth’s threat to invoke the act, however, was described as “incoherent” by AI policymakers, according to a report by POLITICO, raising questions about the strategy’s legal viability and potential ramifications.
Trump characterized Anthropic’s position as a “strong-arm” tactic, stating on Truth Social that the company had “made a mistake.” The President’s order reflects a growing tension between the administration’s desire to leverage AI for national security purposes and the ethical concerns raised by AI developers regarding the potential misuse of their technology.
The dispute highlights the challenges of regulating AI development and deployment, particularly in the context of military applications. Anthropic, led by CEO and co-founder Dario Amodei, has publicly emphasized the importance of responsible AI development and the need to prevent its technology from being used in ways that could harm individuals or society. The company’s reluctance to grant the Pentagon unrestricted access to Claude is rooted in these ethical considerations.
The Pentagon’s pursuit of unrestricted access to Claude underscores its ambition to integrate AI into a wide range of military operations. Officials believe that AI can enhance capabilities in areas such as intelligence gathering, threat analysis, and autonomous weapons systems. However, the disagreement with Anthropic demonstrates the difficulties of balancing these objectives with the need to safeguard against potential risks.
As of Friday evening, Anthropic had not issued a public response to Trump’s order. The Department of Defense has yet to detail its plan for phasing out Anthropic’s technology within the six-month timeframe, and it remains unclear whether the administration will pursue alternative options for securing access to advanced AI capabilities.
