U.S. President Donald Trump has ordered federal agencies to cease using Anthropic’s technology. Following Trump’s directive, the Department of Defense designated Anthropic as a supply chain threat to national security and stated that military contractors and partners must also stop commercial activities. These measures stem from disagreements over Anthropic’s AI model “safety guardrails.”
Last year, Anthropic secured a $200 million AI contract from the Pentagon and planned to apply the technology to defense use cases. However, recent reports indicate that the Department of Defense objected to Anthropic’s model being used for mass surveillance or autonomous weapons development and demanded that the model be available for all lawful purposes.
During this process, Anthropic rejected the Pentagon’s compromise proposal, leading the Department of Defense to classify Anthropic’s models as dangerous. As a result, Anthropic’s technology is expected to be gradually removed from government networks within six months. Anthropic CEO Dario Amodei stated in a blog post that the company will assist in transitioning to other providers to ensure that military plans or operations are not affected.
In related developments, discussions are underway regarding alternative technologies, including collaborations with OpenAI, and reviews of applying other AI companies’ technologies to classified networks.