Chinese AI labs secretly used millions of Claude exchanges to train their models, Anthropic says
- ID
- 23407
- Status
- summarized
- Published
- 11 Sep 2026, 8:59 AM
- Fetched
- 11 Sep 2026, 9:59 AM
- Provider
- CNBC Technology
- Category
- technology
- Original URL
- https://www.cnbc.com/2026/09/11/chinese-ai-labs-moonshot-deepseek-alibaba-anthropic.html
- Source URL
- https://www.cnbc.com/id/19854910/device/rss/rss.html
Summary
- Score
- 7.5
- Created
- 11 Sep 2026, 9:59 AM
- Tags
- Audience
- developersai_ml_learnerssaas_startup_foundersai_agent_users
What happened
Anthropic alleges that Alibaba, Moonshot, and DeepSeek engaged in large-scale unauthorized 'distillation' of Claude outputs to train their own models, with Alibaba using Claude outputs for Qwen and Moonshot routing Kimi user requests through Claude. The findings appear in Anthropic's threat intelligence report covering December 2025 to August 2026, and note that some extracted exchanges contained sensitive data from individual users, multinationals, and state-affiliated actors.
Why it matters
If you ship products on Qwen or Kimi, your users' prompts may have been routed through Claude without disclosure, and the resulting models may carry capabilities distilled from a competitor's proprietary system. Founders evaluating Chinese model providers for cost reasons should weigh this against data-handling transparency and potential legal exposure before committing to a vendor.
Discussion angle
Anthropic is both the victim and the narrator here—how much of this is genuine security reporting versus a competitive move to discourage developers from adopting cheaper Chinese models? What does this mean for Malaysian startups that chose Qwen or DeepSeek for cost reasons?