Anthropic Accuses Alibaba of Distilling Claude

Anthropic told the U.S. Senate Banking Committee and White House officials in a letter, according to Bloomberg and CNBC, that operators linked to Alibaba's Qwen AI lab used nearly 25,000 fraudulent accounts to conduct almost 29 million exchanges with Claude between April 22 and June 5. Per Bloomberg and The Next Web, Anthropic said the exchanges focused on software engineering and agentic reasoning, areas the company describes as among Claude's most commercially valuable capabilities. The Next Web reports Anthropic characterised this as the largest distillation campaign reported to date, exceeding earlier campaigns Anthropic disclosed in February involving DeepSeek, MiniMax, and Moonshot AI that together generated over 16 million exchanges via about 24,000 fake accounts. Alibaba's stock fell approximately 3% - dropping below $100 - on the news per Yahoo Finance.

Distillation is the process of feeding carefully constructed queries to a frontier AI model, collecting its responses, and using those outputs to train a cheaper rival system that approximates the original's capabilities. Industry practitioners and Anthropic's February blog describe the technique as efficient for transferring capabilities, especially for code generation and multi-step reasoning tasks. Large-scale automated querying at the volumes Anthropic reports - tens of millions of exchanges over six weeks - can produce datasets that accelerate a target model's replication of specific skills by capturing edge-case behaviors, multi-step chains, and prompt-response patterns at industrial scale. Anthropic warns that models built via illicit distillation often lack the original lab's safety guardrails, adding a misuse risk beyond the IP concern.

The White House flagged distillation as a national security concern in April, when OSTP Director Michael Kratsios published a memo committing the government to share intelligence with US AI labs about foreign campaigns. Anthropic's letter said the Alibaba operation ran after that memo, in defiance of the administration's stated warnings. Senators Bill Hagerty (R) and Andy Kim (D) plan to introduce a bipartisan amendment to must-pass defence legislation that would blacklist or sanction Chinese firms found improperly accessing US AI model output. A companion House bill backed by Representatives Bill Huizenga and Sydney Kamlager-Dove is also under consideration, though passage of either into the final defence bill is uncertain (The Next Web).

The accusation escalates an already complicated moment for both companies. The Pentagon added Alibaba to its Chinese military companies blacklist on June 8; Alibaba sued the Defense Department this week seeking removal, calling the label baseless. The Anthropic letter opens a second front, framing Alibaba as an active participant in what Anthropic calls systematic theft of American AI capabilities. For Anthropic, the timing is sensitive: the company - valued at $965 billion after a $65 billion Series H - filed confidentially for an IPO this month and is preparing for a potential listing as soon as autumn 2026. US officials have estimated unauthorised distillation costs Silicon Valley labs billions of dollars annually, making it a material risk disclosure for a company heading to public markets. Separately, Anthropic is embroiled in a dispute with the Trump administration over export controls imposed on its Fable 5 and Mythos 5 models, complicating its relationship with the same government it is now asking to act against Alibaba (The Next Web).

Track follow-up reporting and regulatory inquiries citing the letter seen by Bloomberg. Relevant technical signals include anomalous account creation patterns, high-volume low-latency query fingerprints, and concentrated usage on particular capability classes - especially code generation and agentic planning prompts. Watch for industry coordination on provenance, watermarking, or API-level rate-limiting measures, and for any changes to terms of service or access controls mentioned in public filings or regulatory guidance.

This episode highlights a growing operational and policy fault line around large-scale extraction of frontier model behavior. Anthropic has now named four Chinese labs as distillers of its technology, with the Alibaba accusation far exceeding prior campaigns in scale. The naming of a major Chinese conglomerate - rather than smaller AI startups - and the bipartisan legislative response mark a significant escalation in the politics of AI IP protection. Whether the Hagerty-Kim amendment or the House companion bill survive to become law will shape the enforcement landscape for AI model extraction across the industry.