Anthropic letter reveals alleged Alibaba attempt to distill Claude AI amid US restrictions on Mythos and Fable 5 access
On June 12, the US government asked Anthropic to suspend access to its advanced AI models, Mythos and Fable 5. Although the move raised questions at the time, a newly revealed letter from Anthropic sheds light on the concerns that drove these decisions. According to the company, one of China’s largest tech firms may have attempted to access its AI technology on a massive scale.
Ahead of a scheduled hearing on artificial intelligence, Anthropic sent a letter—dated June 10—to Senators Tim Scott and Elizabeth Warren, the Chair and Ranking Member of the US Senate Banking Committee, respectively. In the letter, Anthropic accused Alibaba Group of conducting a large-scale operation to “illicitly” access its Claude AI model using thousands of fraudulent accounts.
According to Anthropic, the campaign took place between April 22 and June 5, 2026, and involved nearly 25,000 fraudulent accounts that generated over 28.8 million interactions with Claude.
The company alleged that this activity was carried out by operators affiliated with Alibaba and Alibaba Qwen, the company’s AI research division. Alibaba had not responded to requests for comment at the time the allegations were made public.
What is AI distillation?
At the heart of the dispute lies a practice known as “distillation.” Simply put, distillation involves training a less capable AI model using responses generated by a more advanced one. Instead of building everything from scratch, developers can learn from a more powerful model’s results and use that information to improve their own systems.
Anthropic maintains that the alleged campaign aimed to accelerate China’s ability to achieve performance levels similar to those of its advanced Mythos Preview model. The company views these efforts as a serious threat, as they could enable competitors to replicate advanced AI capabilities without investing the same level of time and resources in research and development.
Not the first accusation involving Chinese AI companies
This is not the first time Anthropic has raised concerns about model distillation. In February, the company reported identifying another campaign involving the Chinese AI startup DeepSeek, along with two other Chinese labs in the sector. According to Anthropic, DeepSeek’s operation involved more than 150,000 interactions with Claude. The company stated that Moonshot AI was responsible for over 3.4 million interactions, while MiniMax generated more than 13 million.
At the time, Anthropic warned that such campaigns were increasing in both “intensity and sophistication” and maintained that addressing this threat would require swift, coordinated action from the AI industry and governments.
Why the U.S. is paying close attention
Anthropic’s latest accusations come at a time when Washington is showing growing concern over the flow of advanced AI technology. In its letter, the company expressed support for U.S. government efforts to combat these activities through threat information sharing and collaboration with private AI companies.
This month, Alibaba was added to the Pentagon’s list of Chinese military companies—a designation the company is challenging.
At the same time, U.S. officials have adopted a more cautious approach regarding other Chinese AI companies. DeepSeek, for instance, has not been placed on the U.S. trade blacklist, even though an interagency government committee reportedly considers it a national security risk.
Why Mythos and Fable matter
These accusations help explain why access to advanced AI systems like Mythos and Fable has become a sensitive issue.
As AI models become more powerful, companies and governments increasingly view them as strategic assets rather than mere software products. There is a concern that if advanced capabilities can be replicated through large-scale distillation processes, the gap between leading AI developers and their competitors could narrow much faster than anticipated.
For Anthropic, the issue centers on protecting years of research. For policymakers, the matter is becoming part of a much broader debate regarding technology, national security, and the competition between the United States and China.