Anthropic alleges Alibaba extracted Claude AI capabilities in major distillation campaign
By Cygnus | 25 Jun 2026
Summary
Anthropic has accused Alibaba of conducting what it describes as the largest known distillation attack against its Claude AI models. The U.S. AI company claims operators linked to Alibaba’s Qwen AI lab used nearly 25,000 fraudulent accounts to generate 28.8 million interactions with Claude between April and June 2026, allegedly extracting advanced model capabilities to improve competing AI systems. Alibaba has not publicly responded to the allegations.
SAN FRANCISCO, June 24, 2026 — Artificial intelligence company Anthropic has accused Chinese technology giant Alibaba Group of illicitly extracting capabilities from its Claude AI models through a large-scale distillation campaign, according to a letter reviewed by Reuters. Anthropic described the incident as the largest known distillation attack against the company to date.
In a letter dated June 10 and sent to U.S. senators including Tim Scott and Elizabeth Warren, Anthropic alleged that operators affiliated with Alibaba and its Qwen AI lab carried out a coordinated effort to access Claude’s advanced capabilities using nearly 25,000 fraudulent accounts. The company said the campaign generated more than 28.8 million interactions with Claude between April 22 and June 5, 2026.
According to Anthropic, the activity focused on extracting some of Claude’s most valuable capabilities, including software engineering, agentic reasoning and complex long-horizon task execution. The company argues that such distillation efforts allow competitors to accelerate AI development by training models on the outputs of more advanced systems rather than investing in equivalent research and development.
Distillation is a widely used AI training technique in which a smaller or less capable model learns from the outputs of a more advanced model. Anthropic contends that the alleged campaign violated its terms of service and access restrictions, which prohibit commercial access to Claude from China.
The allegations come amid growing concerns in Washington over the protection of advanced U.S. AI technologies and the potential transfer of frontier AI capabilities to foreign competitors. Anthropic urged policymakers to strengthen measures aimed at preventing unauthorized extraction of AI model capabilities and to consider penalties for organizations involved in such activities.
The company said the alleged Alibaba-linked operation exceeded the scale of previously reported distillation campaigns involving Chinese AI firms including DeepSeek, Moonshot AI and MiniMax, which Anthropic disclosed earlier this year.
Alibaba had not publicly responded to requests for comment at the time of publication. Reports indicated that Alibaba shares fell following news of the allegations, reflecting investor concerns over potential regulatory and geopolitical repercussions.
Why this matters
- AI intellectual property protection: The case highlights growing concerns about protecting proprietary AI capabilities as competition among global AI developers intensifies.
- Geopolitical AI rivalry: The allegations add to tensions between the United States and China over access to advanced AI technologies and computing resources.
- Model distillation scrutiny: The incident could increase regulatory attention on how AI companies use model outputs to train competing systems.
- Market impact: Investor reaction shows that AI-related intellectual property disputes can have immediate consequences for major technology companies.
FAQs
Q1: What is Anthropic accusing Alibaba of?
Anthropic alleges that operators linked to Alibaba’s Qwen AI lab used nearly 25,000 fraudulent accounts to generate 28.8 million interactions with Claude AI models and extract advanced capabilities through a large-scale distillation campaign.
Q2: What is AI distillation?
Distillation is a machine learning technique in which a less capable model is trained using the outputs of a more advanced model to improve performance.
Q3: Which Claude capabilities were allegedly targeted?
Anthropic said the campaign focused on software engineering, agentic reasoning and complex long-horizon task capabilities.
Q4: How many interactions were involved in the alleged campaign?
Anthropic claims the operation generated more than 28.8 million interactions with Claude between April 22 and June 5, 2026.
Q5: Has Alibaba responded to the allegations?
As of publication, Alibaba had not publicly responded to the claims.