Search
On February 24 (Beijing Time), U.S.-based AI firm Anthropic alleged that Chinese developers, including DeepSeek, Moonshot AI, and MiniMax, had created over 24,000 fake accounts. These accounts were used to interact with Anthropic's AI model, Claude, more than 16 million times. According to Anthropic, these developers employed 'distillation' techniques—a method where knowledge from a larger model is transferred to a smaller one—to essentially steal its model's capabilities and enhance their own AI models. In a counter-move, Elon Musk criticized Anthropic, claiming that the company itself had previously engaged in large-scale theft of training data, leading to settlements amounting to billions of dollars. Musk accused Anthropic of hypocrisy, pointing out the apparent double standards in their accusations.
-
C114 Communication Network -
Communication Home
