Pentagon pressures Anthropic on AI guardrails
The U.S. Department of Defense is in a high-stakes standoff with AI developer Anthropic, which is valued at $380 billion. The Pentagon is pushing the company to relax its AI safety guardrails for military applications. The negotiations could set a major precedent for the use of artificial intelligence in warfare and the boundaries between civilian tech companies and military deployment.
Anthropic, a company founded in 2021 by former OpenAI employees, has positioned itself as a leader in AI safety. This safety-first approach was a key differentiator for the company, which recently saw its valuation surge to $380 billion after a $30 billion funding round in February 2026. The conflict with the Pentagon centers on Anthropic's refusal to remove two key "red lines" from its contracts. The company, led by CEO Dario Amodei, will not permit its AI model, Claude, to be used for mass domestic surveillance or in fully autonomous weapons systems that can kill without human intervention. The Department of Defense, under the leadership of Defense Secretary Pete Hegseth, is demanding "all lawful use" of the technology. Pentagon officials argue that it is their role, not a private company's, to determine the legal and ethical boundaries of military operations. The Pentagon's Chief Technology Officer, Emil Michael, has been a key figure in these negotiations. This standoff has significant consequences for Anthropic, which holds a $200 million contract with the DoD and is the only AI company with its model deployed on the Pentagon's classified networks. The Pentagon has threatened to cancel the contract and designate Anthropic a "supply chain risk," a move that could have severe financial implications. The dispute highlights a broader struggle between Silicon Valley and the U.S. government over the ethical boundaries of AI in warfare. While Anthropic has held firm, other major AI players like Google, OpenAI, and Elon Musk's xAI have also secured military contracts, with xAI reportedly agreeing to the Pentagon's "all lawful use" framework. Anthropic's technology has already been utilized in sensitive military operations, including the capture of Venezuelan leader Nicolás Maduro. This prior use has intensified the debate, with Anthropic reportedly raising questions about how Claude was used in that raid, further straining the relationship. Recently, Anthropic revised its own Responsible Scaling Policy, removing a provision that required the company to halt the training of more powerful models if safety capabilities couldn't keep pace. The company stated this change was due to a policy environment prioritizing AI competitiveness over safety discussions. The deadline for Anthropic to comply with the Pentagon's demands was set for Friday, February 27, 2026. CEO Dario Amodei has publicly stated that the company "cannot in good conscience" agree to the Pentagon's terms, setting the stage for a potentially groundbreaking decision on the future of AI in national security.
