The Pentagon is actively testing competing artificial intelligence models as it looks for alternatives to Anthropic’s Claude system, which has become deeply integrated into several U.S. military operations. According to reports, the Department of Defense has been evaluating AI systems from companies such as OpenAI and Google through a group of approximately 25 military “power users” spread across multiple global combat commands. The effort follows growing tensions between the Pentagon and Anthropic over restrictions and safety guardrails attached to the company’s AI technology.
The conflict began after Defense Secretary Pete Hegseth reportedly labeled Anthropic a potential supply-chain risk because the company insisted on maintaining limitations around how its AI could be used in military operations. Anthropic has argued that certain safeguards are necessary to prevent its systems from being deployed in fully autonomous weapons or domestic surveillance activities. The disagreement escalated into legal battles, with Anthropic claiming that the Pentagon’s designation could seriously damage its business and future government partnerships.
Despite the dispute, the U.S. military has already relied heavily on Claude within Palantir’s Maven Smart System, a digital mission platform reportedly used in classified operations involving Iran and other international conflicts. Military officials have praised Claude for its usability and performance, but the Pentagon now appears determined to diversify suppliers and avoid depending on a single AI provider. New agreements with rival companies are being developed for deployment on classified government networks as defense agencies accelerate adoption of military AI tools.
The broader geopolitical and ethical tensions surrounding AI in warfare. Supporters argue that advanced AI systems can improve intelligence analysis, decision-making speed, and operational coordination on the battlefield. However, human rights organizations and AI safety advocates warn that rapidly integrating AI into military systems increases risks related to automation, accountability, and unintended escalation during conflicts. The Pentagon’s search for replacement models reflects a wider global competition among governments and technology companies to control the future of AI-powered defense infrastructure.
