U.S. Defense Secretary Pete Hegseth delivered an ultimatum to artificial intelligence company Anthropic during a meeting this week with its chief executive, escalating a months-long dispute over the military use of advanced AI systems, according to people familiar with the matter. During the meeting with Anthropic CEO Dario Amodei, Hegseth set a deadline of Friday following the talks for the company to agree to terms allowing broader military use of its technology.
Pentagon officials want Anthropic to comply. They warned that if Anthropic does not comply, the government could take drastic steps. It includes designating the company as a supply-chain risk or invoking the Defense Production Act, which would allow the government to compel cooperation in the name of national security. Anthropic has resisted removing safeguards that restrict its AI models from being used for autonomous weapons targeting or mass domestic surveillance to ensure responsible deployment.
The dispute comes as the Department of Defense accelerates efforts to integrate artificial intelligence across battlefield operations, cyber defense, and intelligence analysis. The Pentagon is negotiating contracts with multiple large language model providers, including xAI and OpenAI after announcing this week that xAI had received approval to deploy its models on classified networks. Until recently, Anthropic was the only AI firm cleared for such access.
Anthropic said the meeting with defense officials reflected ongoing discussions and that it remains committed to supporting the government’s national security mission “in line with what our models can reliably and responsibly do.”
Founded in 2021 by former OpenAI researchers and executives, Anthropic has positioned itself as a leading advocate for AI safety. The company develops the Claude family of large language models and has built a sizable enterprise business, with more than 500 customers spending over $1 million annually. Anthropic recently closed a $30 billion funding round valuing the company at $380 billion, underscoring its growing influence as debates intensify over how powerful AI should be used in national security.
Also Read: Anthropic’s Dario Amodei on the hotseat with Pentagon over Military AI
