News
Public · Published
China Probes DeepSeek and Moonshot Over Alleged Data Leaks to Anthropic's Claude
China's internet regulator is investigating DeepSeek and Moonshot AI. Anthropic accused both companies of secretly routing millions of user exchanges through Claude, its AI model, to train their own models.
Published:
Updated:
What happened
China's internet regulator is investigating DeepSeek and Moonshot AI. Anthropic accused both companies of secretly routing millions of user exchanges through Claude, its AI model, to train their own models.
Confirmed
Global impact / market context
If confirmed, the investigation could raise legal and compliance costs for DeepSeek and Moonshot, which are Chinese AI companies. It may also chill data-sharing deals between AI developers and foreign model providers, slowing model training.
Analyst inference
The probe highlights growing tension over how AI companies obtain training data. Investors in Chinese AI firms may reprice regulatory risk, while providers like Anthropic could tighten access controls, affecting partnerships and cross-border data flows.
Analyst inference
What to watch
- Watch whether China's internet regulator publicly confirms the investigation and names the specific data-protection or cybersecurity rules DeepSeek and Moonshot are alleged to have broken. Confirmed
- Anthropic could restrict Claude access for Chinese developers or add technical controls to prevent bulk data extraction, which means limits on how much user data can be sent. Proposed
- If regulators demand audits or fines, DeepSeek and Moonshot may face higher costs and slower model releases, potentially hurting their ability to compete with global AI rivals. Analyst inference