Anthropic and the Pentagon are reportedly arguing over Claude usage

Recent reports indicate a growing conflict between the Pentagon and AI firm Anthropic regarding the use of its technology. The U.S. military is advocating for AI companies to permit their systems to be utilized for 'all lawful purposes.' However, Anthropic appears to be resisting these demands, as highlighted in a report by Axios. The Pentagon has also reached out to other major AI companies, including OpenAI, Google, and xAI, with similar requests. An anonymous source from the Trump administration disclosed that one of these companies has acquiesced, while the others are reportedly open to negotiation. In contrast, Anthropic has been described as the most reluctant participant in these discussions. As tensions escalate, the Pentagon is reportedly contemplating the termination of its substantial $200 million contract with Anthropic. This follows earlier reports from the Wall Street Journal, which indicated notable disagreements between Anthropic and Defense Department officials regarding the application of its Claude models. The WSJ also mentioned that Claude was utilized in a military operation aimed at apprehending Nicolás Maduro, the former President of Venezuela. In response to inquiries from TechCrunch, Anthropic did not provide immediate comments. However, a company spokesperson stated to Axios that discussions regarding Claude's application for specific military operations have not taken place. Instead, Anthropic is concentrating on defining its Usage Policy, particularly concerning its strict prohibitions against fully autonomous weapons and extensive domestic surveillance.