A significant conflict has unfolded in the United States over who should control the application of powerful generative AI models in military contexts. During contract negotiations, the U.S. Department of Defense (Pentagon) demanded more extensive, effectively unlimited, access to Anthropic’s AI model, Claude.