China Warns Claude Code AI Tool Contains Backdoor Sending User Data to Remote Servers; Anthropic Disputes Claims
Summary
China's Ministry of Industry and Information Technology warns that Anthropic's Claude Code AI tool contains a backdoor secretly transmitting sensitive user data — including location and identity — to remote servers, affecting versions released between April and June, while Anthropic disputes the claim, calling it an anti-distillation experiment amid escalating U.S.-China AI tensions.
Key Points
- China's Ministry of Industry and Information Technology is warning that Anthropic's Claude Code AI coding tool contains a 'back-door' security vulnerability capable of sending sensitive user data, including location and identity, to a remote server without consent.
- Affected versions of Claude Code span 2.1.91 to 2.1.196, released between April 2 and June 29, and users are being urged to uninstall or upgrade immediately.
- Anthropic is firing back, stating the 'backdoor' was an anti-distillation experiment and reminding users that its policy prohibits use by entities majority-owned by China-headquartered organizations, amid rising U.S.-China AI tensions.