China warns of security risks in relay services for foreign AI models
China's Ministry of Security has raised alarms regarding the dangers associated with grey-market relay services that enable access to foreign AI models, such as Anthropic's Claude and the GPT series from Microsoft-supported OpenAI, which are unavailable in mainland China. The ministry advised developers that certain relay platforms store user data on their servers without sufficient encryption, posing threats of data breaches or unauthorized trading. This alert arrives as an increasing number of Chinese developers seek these services to utilize US models with superior coding abilities, even as Chinese AI startups make progress.
Key facts
- China's security ministry warns of security risks in relay services for foreign AI models.
- Relay services provide access to Anthropic's Claude and OpenAI's GPT series.
- These AI models do not officially support users in mainland China.
- Some relay platforms retain user data without adequate encryption.
- Risks include data leaks and illicit trading of sensitive information.
- Chinese developers seek US models for stronger coding capabilities.
- Chinese AI startups have made advances but are still sought after for coding.
- The warning was issued by the security ministry.
Entities
Institutions
- Anthropic
- Microsoft
- OpenAI
- Shutterstock
Locations
- China
- mainland China