Anthropic CEO Dario Amodei could still be trying to make a deal with Pentagon

Anthropic CEO Dario Amodei has unexpectedly resumed negotiations with Pentagon official Emil Michael, according to new reports from the Financial Times and Bloomberg, just one week after their $200 million Department of Defense contract collapsed. The original deal broke down when Amodei refused to accept a clause granting the military unrestricted "any lawful use" access to Anthropic's AI models, specifically citing concerns about potential domestic mass surveillance and autonomous weaponry applications. When Anthropic stood firm, the Pentagon pivoted to strike a deal with OpenAI instead, leading to a public feud where Michael called Amodei a "liar" with a "God complex," while Amodei dismissed OpenAI's agreement as "safety theater" and accused them of "straight up lies."

Despite the vitriolic exchange, both sides have practical reasons to compromise—the Pentagon already relies on Anthropic's technology infrastructure, and abruptly switching to OpenAI's systems would be operationally disruptive. The renewed talks reportedly focus on crafting a new contract with clearer prohibitions against specific military applications that violate Anthropic's safety principles. Meanwhile, Defense Secretary Pete Hegseth has threatened to declare Anthropic a "supply-chain risk," which would effectively blacklist the company from working with any U.S. military contractors, though this unprecedented move against a domestic AI firm would likely face legal challenges. The outcome will set crucial precedents for how AI companies negotiate ethical boundaries with government agencies while maintaining access to major contracts.