Anthropic Accuses Alibaba of AI Theft

Anthropic has accused Alibaba Group of orchestrating the largest known 'distillation attack' on its Claude AI model. This alleged campaign involved 28.8 million exchanges with Claude through almost 25,000 fraudulent accounts between April and June, aiming to extract Claude's capabilities for Alibaba Group's Qwen App AI lab. Anthropic claims this practice, known as adversarial distillation, allows Chinese companies to develop rival chatbots at a fraction of the cost, undermining US AI developers and posing a threat to national security. Anthropic has sent a letter to US senators and White House officials, urging stronger government action, including clarifying antitrust guidelines, tightening chip controls, and penalizing firms using distillation. This event adds to existing political pressure on Alibaba Group, which was recently added to a US Defense Department blacklist and is currently suing the Pentagon to be removed. The US government has also taken steps to address AI intellectual property theft, with a White House memo from Michael Kratsios and proposed legislation from Senators Bill Hagerty and Andy Kim. However, the United States — United States Department of Commerce also imposed export controls on Anthropic's Fable 5 and Mythos 5 AI systems, citing security concerns, creating a dispute with Anthropic.