The Pentagon has raised concerns about the potential risks associated with granting Anthropic access to its warfighting systems, labeling it as an "unacceptable risk" to national security. This assertion comes from a court filing responding to the AI firm's legal challenge against the government's supply chain risk designation. Anthropic's refusal to comply with terms set forth in AI service contracts prompted the Department of Defense to question the company's status as a "trusted partner" for sensitive operations.
In the filing, the Department of War, as referred to by the Trump administration, emphasized that AI technologies are particularly susceptible to manipulation. The Pentagon indicated that Anthropic could modify or disable its systems during critical operations if its corporate policies were infringed upon. Due to these vulnerabilities, former President Trump instructed federal agencies to cease using Anthropic's technology.
As a result of its legal action, Anthropic is seeking a preliminary injunction to halt the ban while contesting its supply chain classification. The firm claims this designation could lead to significant financial losses, potentially amounting to billions in revenue, as it continues to work with clients on non-defense projects. Meanwhile, major tech companies like Microsoft, Google, and OpenAI have shown support for Anthropic by submitting friend-of-the-court briefs related to the case.