Chinese AI Firms Launch Large-Scale Distillation Campaign Against Anthropic's Claude
AI company Anthropic claims that three Chinese artificial intelligence firms—DeepSeek, Moonshot and MiniMax—created 24,000 fake accounts and conducted 16 million interactions with Claude to steal training data.
AI company Anthropic has leveled allegations against three artificial intelligence companies for illegally exploiting its Claude large language model to enhance their own AI systems through what is referred to as a "distillation" attack.
Through a blog post published on Sunday, Anthropic revealed that it had uncovered these "attacks" carried out by DeepSeek, Moonshot, and MiniMax, a technique that involves using outputs from a more advanced model to train an inferior one.
The company claimed that the three entities collectively created "over 16 million exchanges" with Anthropic's Claude AI system using "approximately 24,000 fraudulent accounts."
"Distillation is a widely used and legitimate training method. For example, frontier AI labs routinely distill their own models to create smaller, cheaper versions for their customers," Anthropic wrote, adding:
"But distillation can also be used for illicit purposes: competitors can use it to acquire powerful capabilities from other labs in a fraction of the time, and at a fraction of the cost, that it would take to develop them independently."
According to Anthropic, the campaigns were designed to extract capabilities from Claude across multiple domains, including agentic reasoning, coding and data analysis, rubric-based grading tasks, and computer vision.
"Each campaign targeted Claude's most differentiated capabilities: agentic reasoning, tool use, and coding," stated the multi-billion-dollar artificial intelligence company.
The company stated it successfully identified the three organizations through "IP address correlation, request metadata, infrastructure indicators, and in some cases corroboration from industry partners who observed the same actors and behaviors on their platforms."
All three companies—DeepSeek, Moonshot, and Minimax—operate out of China as AI firms. The trio carries estimated valuations reaching into the multi-billion dollar territory, with DeepSeek maintaining the highest level of international brand recognition among them.
In addition to concerns about intellectual property theft, Anthropic contended that such distillation operations conducted by international competitors pose serious geopolitical security threats.
"Foreign labs that distill American models can then feed these unprotected capabilities into military, intelligence, and surveillance systems—enabling authoritarian governments to deploy frontier AI for offensive cyber operations, disinformation campaigns, and mass surveillance," the firm said.
Looking ahead, Anthropic announced it would implement defensive measures by improving detection systems to identify suspicious traffic patterns, sharing threat intelligence, and implementing stricter access controls, among other protective strategies.
The company further advocated for increased cooperation from domestic industry participants and lawmakers to help stop foreign AI companies from attacking US firms.
"No company can solve this alone. As we noted above, distillation attacks at this scale require a coordinated response across the AI industry, cloud providers, and policymakers. We are publishing this to make the evidence available to everyone with a stake in the outcome."