U.S. Department of Defense halts use of Anthropic AI

The U.S. Department of Defense has completely ceased the use of Anthropic’s ‘Claude’ AI models, following a dispute over safety protocols. The decision comes eight months after Defense Secretary Pete Hegseth designated Anthropic as a ‘supply chain risk company’ after the firm refused requests to remove safety guardrails that would allow unrestricted military use, including potential applications in mass surveillance and autonomous weaponry.

Removing Claude proved difficult due to its deep integration into the ‘Maven Smart System,’ a large-scale data platform operated by Palantir used for military analysis and decision-making. The transition period was extended because the technology was extensively used for research, intelligence gathering, and information classification during recent conflicts, including the war in Iran.

As Anthropic is phased out, the Department of Defense is expanding its contracts with other major AI developers, including OpenAI, Google, and xAI. This shift occurs amid broader industry debates regarding the speed of artificial intelligence development. While some industry leaders and former researchers warn of existential risks from unaligned Artificial General Intelligence (AGI), others suggest these warnings may serve as marketing strategies to establish control over safety standards.