Anthropic CEO Dario Amodei is negotiating with U.S. Under Secretary of Defense for Research and Engineering Emil Michael to try to reach a compromise on the terms of a Pentagon contract to use Anthropic's AI model. Negotiations broke down last week. In an internal memo, Amodei stated that at the end of the negotiations, the Department of Defense proposed removing a clause from the contract regarding "bulk acquisition and analysis of data," a core safeguard preventing Anthropic from using AI for mass domestic surveillance. Anthropic's red lines include prohibiting the use of AI for mass domestic surveillance and lethal autonomous weapons. Amodei also stated in the memo that some statements from the Pentagon and OpenAI contained misinformation and implied that Anthropic was sidelined because it did not praise Trump. Previously, U.S. Defense Secretary Pete Hegseth threatened to classify Anthropic as a supply chain risk, a move that would force companies in the military supply chain to sever ties with Anthropic. Anthropic won a $200 million contract from the U.S. Department of Defense last July, becoming the first AI model to be used in classified environments and by national security agencies. (FT)