Anthropic accuses Alibaba of large-scale AI model extraction campaign

U.S.-based artificial intelligence company Anthropic has accused Chinese technology giant Alibaba of conducting what it describes as the largest known attempt to extract capabilities from its Claude AI models, according to a report by Reuters.The allegations centre on a practice known as "distillation", in which a less advanced AI model is trained using outputs generated by a more sophisticated system. Anthropic claims the activity was aimed at reproducing some of the capabilities of its latest AI models.According to the letter, the campaign took place between April 22 and June 5, 2026, and involved more than 28.8 million interactions with Claude through nearly 25,000 accounts that Anthropic described as fraudulent.The company alleged that the effort was carried out by operators affiliated with Alibaba and its AI research division, Qwen. Alibaba had not publicly responded to the claims at the time of reporting.Anthropic said the scale of the activity far exceeded previous incidents it had identified involving other Chinese AI developers. Earlier this year, the company disclosed similar allegations involving Chinese startups including DeepSeek, Moonshot AI and MiniMax, which it said had collectively generated millions of interactions with Claude in attempts to extract model capabilities.The allegations were outlined in a letter dated June 10 and sent to members of the U.S. Senate Banking Committee ahead of a hearing focused on artificial intelligence and national competitiveness.The dispute comes amid growing concerns in Washington over the protection of advanced AI technologies and intellectual property. U.S. officials have increasingly warned about efforts to acquire cutting-edge AI capabilities developed by American companies.Anthropic said it supports government efforts to address such threats through information-sharing partnerships between AI companies and public agencies.The claims also emerge against a backdrop of rising geopolitical tensions surrounding artificial intelligence. Earlier this month, Alibaba was added to a Pentagon list identifying companies with alleged links to China's military, a designation the company is contesting.Separately, U.S. authorities recently imposed restrictions on some of Anthropic's latest AI models over concerns that advanced systems could be accessed by military or intelligence users in countries considered sensitive by Washington.Also read: Claude Code creator Boris Cherny flags risks of AI writing 100% of code, calls it ‘problematic’