You thought you were talking to DeepSeek or Kimi? Surprise! It was often Claude from Anthropic, behind the scenes. This revelation unveils a dangerous game between AI giants.
Claude from Anthropic: the discreet star hijacked by its Chinese competitors
Claude, Anthropic’s AI model, has become the star of a strange industrial soap opera. Since December, seven Chinese laboratories, including Alibaba, DeepSeek, Moonshot AI, and Xiaomi, are said to have hijacked nearly 200 million interactions, according to an explosive report from Anthropic published in early September. This industrial siphoning operation does not resemble a mere classical imitation.
The technique used is called “distillation”: a smaller model learns to respond based on the answers from a larger model, the “teacher.” It is a standard method, but here, the “teacher” did not consent! Thus, the Chinese laboratories are said to have made Claude respond in place of their own services, subsequently displaying the answers as their own creations.
For instance, Moonshot AI transferred 300,000 customer requests to Claude via more than 5,000 fraudulent accounts based in Singapore and Japan. DeepSeek, on the other hand, developed a more sophisticated mechanism. It captured requests from developers using tools like Claude Code and secretly redirected them to Claude Opus.
The massive data looting at the heart of an industrial chain
Alibaba does not do things by halves, with a peak of 151 million intercepted exchanges in just three months. This represents up to 3 million interactions per day, orchestrated through more than 3,500 fake accounts, using stolen credit cards and fake identities.
Zhipu and Xiaomi complete this chaotic picture by reinjecting user sessions into Claude, while SenseTime sources transcripts through resellers. MiniMax, for its part, operates through a shell company selling only American models, with no Chinese model… a subtle strategy.
This true cascade trafficking showcases the extent of the logistics chain for AI theft. Intermediaries seize conversations, extract reasoning from them, and then feed their own systems.
Impact and issues beyond rivalries between laboratories
This conflict is no longer limited to a mere game of dupes among competitors. Since 2025, OpenAI and Google have also pointed out this phenomenon of illicit distillation. What changes this time are the evidences of an organized system overseeing API theft and credit card fraud. Anthropic thus speaks of “practices likely incompatible with GDPR,” as sensitive data circulates through European platforms unchecked.
The risks extend beyond mere “copy-paste.” The report reveals sensitive data, including email addresses, still valid access tokens, and even investment forecasts from a pharmaceutical laboratory. More shockingly: a user, seemingly linked to the Chinese military, allegedly sent surveillance videos for analysis, thinking they were chatting with Kimi, when they were actually facing Claude.
For advocates of transparency, this case revives the urgency to strengthen digital sovereignty and protect the integrity of AI models. Trust no longer stems from mere statements but from the transparency of mechanisms and open control of data.
Innovative protection methods by Anthropic against data theft
In the face of these attacks, Anthropic has revised its strategy and implemented robust measures. The Claude model now incorporates a reasoning summary system to reduce the value of the stolen data. Furthermore, invisible marking solutions are being studied to trace and authenticate generated texts, thus limiting abusive reuse.
This fight for digital trust is not isolated. Many experts are interested in this invisible marking — a promising avenue for authenticating and securing generated content.
You can consult detailed information on these innovations on this page dedicated to Anthropic’s traceability solutions or on emerging digital marking methods at industry leaders.
List of key points revealed by Anthropic’s report on DeepSeek and Kimi
- Over 200 million hijacked exchanges, targeting seven Chinese laboratories including Alibaba and Xiaomi.
- Massive use of fraudulent accounts with identity theft and credit cards.
- Secret redirection of user requests to Claude, without users knowing.
- Theft of sensitive information including military and industrial data.
- Illegal logistical chain of intermediaries reselling access to Claude at scale.
To dive deeper into this complex affair
The complete report from Anthropic is accessible, but for a comprehensive summary, you can read a clear and critical dossier on the wild distillation of Chinese AIs. A fascinating and worrying context that free digital cannot ignore.
Another major resource, quantumzeitgeist.com details with precision the hijacking techniques used in this case.
This video provides a comprehensive overview of the scale and methods of the hijacking.
Watch this analysis to understand the distillation of AI from a technical perspective and its implications for digital sovereignty.
What is ‘distillation’ in AI?
Distillation is a training method where a smaller model learns to reproduce the responses of a larger model, effectively reducing computational load.
Why does this hijacking pose a problem?
Because it exposes unprotected sensitive data and steals the work of an AI model without permission, deepening the gap of unfair practices and raising confidentiality issues.
How is Anthropic trying to protect itself?
Anthropic has implemented strengthened filters, a reasoning summary of Claude and is developing an invisible marking system to authenticate its content.
Are users aware?
Often, users do not realize they are ultimately interacting with Claude while thinking they are using DeepSeek or Kimi.
Could this scandal have a legal impact?
Yes, as it touches on intellectual property, personal data, and questions GDPR compliance, particularly concerning data circulating via European platforms.
Source: www.01net.com