Hegseth warns Anthropic to let the military use the company’s AI tech as it sees fit, AP source says

Pete Hegseth, a senior advisor to the Secretary of Defense, has issued a warning to AI company Anthropic, urging it to allow the U.S. military to use its AI models, like Claude 3 Opus and Claude 3.5 Sonnet, without restrictive ethical guardrails, according to an Associated Press source. The pressure reflects the Pentagon's aggressive push to operationalize cutting-edge artificial intelligence for tasks ranging from logistics and cyber defense to battlefield planning and intelligence analysis, placing it on a potential collision course with tech firms that have built usage policies to prevent their systems from being weaponized or causing harm.

The core of the dispute centers on Anthropic's Constitutional AI framework, a set of safety principles designed to align its models with human values, which may inherently limit certain military applications. Hegseth's argument, representing a faction within defense circles, is that national security needs must override corporate ethical constraints, especially against near-peer adversaries like China who are not similarly restrained. This confrontation signals a critical inflection point for the AI industry, forcing companies to choose between lucrative government contracts and their publicly stated commitments to responsible AI development, with implications for future model training, deployment, and the very governance of powerful AI systems.