US Accuses Chinese AI Firms Including Alibaba of Proprietary Model Theft

Key Facts

1US cybersecurity and intelligence agencies accused six China-based AI firms, including Alibaba, of industrial-scale extraction of proprietary features from US models.

2Targeted models included variants of Claude, GPT, Gemini, and Grok, with billions of tokens extracted via millions of requests.

Amid escalating technological competition, US cybersecurity and intelligence agencies have accused six China-based AI firms of industrial-scale extraction of proprietary features from American AI models. According to reports, the targeted systems included variants of Claude, GPT, Gemini, and Grok, with billions of tokens extracted via millions of requests. Federal agencies, including CISA and the FBI, allege that these firms utilized proxy accounts and transfer stations to bypass geographic blocks and harvest synthetic training data from leading US systems.

These allegations place fresh pressure on the Chinese tech sector, specifically naming major entities such as Alibaba, DeepSeek, and Moonshot AI. The agencies contend that these firms employed "knowledge distillation" methods to close the technological gap with the US and accelerate the development of domestic models like Qwen and Kimi. Per market data, such geopolitical friction increases regulatory and reputational risks for Chinese firms that are heavily invested in AI infrastructure and international expansion.

Regarding market performance, Alibaba (BABAF) stood at $13.76 at the close of September 10, 2026, following a session high of $16 and a low of $12.74. Investors are closely monitoring the fallout of these accusations on broader trade relations, particularly following China's Balance of Trade data released on September 8, 2026, which showed a surplus of $119.1 billion, highlighting the significant economic stakes involved in this ongoing tech dispute.