Anthropic Accuses Chinese AI Firms of Cloning Models via User Queries

Key Facts

Amid the intensifying global race for AI supremacy, new challenges regarding the intellectual property of large language models are emerging. Anthropic has accused Chinese AI firms DeepSeek and Moonshot of utilizing millions of real user queries to clone its proprietary model capabilities. According to reports, the Chinese entities allegedly deployed thousands of fake accounts to funnel queries into U.S. models, aiming to extract advanced capabilities through a process known as model distillation.

These allegations deepen the technical and geopolitical divide between Washington and Beijing in the artificial intelligence sector, as such practices are viewed as attempts to bypass development hurdles and accelerate competitive parity. Per market data, direct pricing for the involved private entities is unavailable; however, the dispute reflects mounting pressure on the broader tech sector, which faces heightened scrutiny over data security and intellectual property rights.

Looking ahead, traders are closely monitoring potential regulatory responses that could impact listed big-tech counterparts. With instrument prices unavailable as of the September 10, 2026 close, focus remains on evolving policy shifts. Additionally, economic data showed China's Balance of Trade reached 119.1 billion on September 8, 2026, a factor that may influence the broader trade and technology dialogue between the two superpowers.