02/03/2026
A major AI policy clash is unfolding inside the U.S. government - involving the Pentagon, Anthropic, OpenAI, and direct action from President Trump.
President Trump has ordered all U.S. federal agencies to immediately stop using Anthropic’s AI systems, including Claude models. The decision followed a Pentagon classification labeling Anthropic as a “supply chain risk.” Reports indicate the Department of Defense requested Anthropic to remove certain safety safeguards - including restrictions against fully autonomous weapons systems and large-scale domestic surveillance of American citizens.
Anthropic’s leadership refused the request, stating they “cannot in good conscience” agree to loosen those protections. The company plans to challenge the government’s designation in court.
At the same time, OpenAI moved quickly to secure an agreement with the Pentagon. The deal allows OpenAI’s AI models to be deployed on classified U.S. Department of Defense networks. OpenAI clarified that its technology may be used for all lawful government purposes, while maintaining defined red-line policies around harmful misuse.
This development highlights a growing divide over AI governance, national security priorities, military applications of frontier AI models, and the balance between innovation and safety. The outcome could shape how governments worldwide integrate advanced AI into defense infrastructure.