Anthropic flags massive unauthorized Claude distillation by Chinese AI labs
TL;DR Summary
Anthropic says Alibaba, Moonshot, DeepSeek and other China-based labs conducted large-scale, unauthorized distillation of its Claude model to train their own systems, routing millions of exchanges and creating fraudulent accounts. The activity, spanning December 2025 to August 2026, included Alibaba’s operation with over 151 million Claude exchanges (May–July) and Moonshot/DeepSeek forwarding user requests to Claude to train their models, raising privacy and terms-of-service concerns and citing potential violations of laws.
- Chinese AI labs secretly used millions of Claude exchanges to train their models, Anthropic says CNBC
- How Chinese AI Firms Tried to Clone U.S. AI Models WSJ
- Anthropic claims Moonshot, DeepSeek secretly diverted user requests to Claude South China Morning Post
- Anthropic details distillation campaigns from Alibaba, Moonshot AI, and DeepSeek TechCrunch
- China ‘Willing’ to Hold AI Talks With US Despite Row Over Models Bloomberg.com
Reading Insights
Total Reads
0
Unique Readers
0
Time Saved
2 min
vs 3 min read
Condensed
88%
565 → 70 words
Want the full story? Read the original article
Read on CNBC