Anthropic vs. the U.S. Department of Defense: Why "all lawful uses" is a dangerous fiction

The dispute between Anthropic and the U.S. Department of Defense over whether the government can require AI companies to open up "all lawful uses" has become public. The two sides have not reached agreement on mass surveillance of American citizens and lethal autonomous weapons not under human control. The Department of Defense has listed Anthropic as a "supply chain risk," and the company is contesting this in court. The civilian AI contract rules drafted by the U.S. General Services Administration also propose adopting this standard. Commentators argue for replacing it with a unified reference to international human rights law and international humanitarian law.