Anthropic, a prominent American artificial intelligence firm, has publicly accused two Chinese AI developers, DeepSeek and Moonshot, of unauthorized use of its technology. According to Anthropic, these companies have been covertly redirecting user requests to the Claude AI model. The allegations suggest that when users interact with the Chinese platforms, the systems behind the scenes send those prompts to Claude, receive the generated response, and then display that response to the user as if it were produced by the Chinese firms' own internal models.

This accusation represents a significant escalation in the ongoing tensions regarding AI development between the United States and China. The claim follows a week of heightened scrutiny from U.S. officials, as multiple government agencies—including the FBI and the NSA—have recently issued warnings or statements regarding the practices of China’s AI industry. While Anthropic has framed this as a direct violation of its service terms and intellectual property, the companies involved, DeepSeek and Moonshot, have not yet provided a detailed public rebuttal to these specific claims. The situation highlights growing concerns over data integrity and the competitive practices within the global AI market, where companies are racing to demonstrate superior model performance while navigating complex international regulations.