Anthropic Accuses Chinese Companies of Siphoning Data From Claude

Slashdot

by msmash

February 23, 2026

AI-Generated Deep Dive Summary

U.S.-based AI startup Anthropic has accused three Chinese artificial intelligence companies—DeepSeek, Moonshot AI, and MiniMax—of creating over 24,000 fraudulent accounts to access its Claude AI model. In a blog post, Anthropic revealed that these companies exploited the system more than 16 million times, siphoning data to train and enhance their own products through a technique called distillation. While distillation has legitimate uses, such as building smaller versions of existing models, Anthropic claims it was used here to rapidly develop competitive AI products at a fraction of the time and cost. The practice involves extracting knowledge from Claude to improve other systems, which Anthropic says violates their terms of service. DeepSeek, one of the companies involved, had 150,000 interactions with Claude, while Moonshot AI and MiniMax engaged in over 3.4 million and 13 million interactions, respectively. This incident comes after OpenAI, another major AI company, accused DeepSeek of using similar tactics to mimic its own products in a memo sent to House lawmakers earlier this month. The issue highlights the ethical and legal challenges surrounding AI development and data usage. While distillation can be used legitimately, Anthropic warns that misuse could undermine innovation and fair competition in the AI industry. The allegations also raise concerns about data security and intellectual property rights in an increasingly competitive global AI landscape. This matter is particularly significant as it underscores the need for clearer guidelines and oversight to prevent such practices. For tech enthusiasts and professionals, this story sheds light on the potential risks of AI model exploitation and the importance of ethical practices in the field. It also reflects broader tensions between U.S. and Chinese AI companies, with implications for international competition and collaboration in artificial intelligence research. As investigations continue, the case will likely influence discussions around AI governance and innovation policies worldwide.

Verticals

tech

Originally published on Slashdot on 2/23/2026