China Warns of Security Risks in Anthropic's AI Coding Tool
Summary
China's National Vulnerability Database has warned of security risks in Anthropic's AI coding tool 'Cloud Code', advising users to test for vulnerabilities or uninstall it. Alibaba will ban the tool from July 10 due to these concerns.
Key Points
- China's National Vulnerability Database has identified a security backdoor in Anthropic's AI coding tool 'Cloud Code' that leaks sensitive user data.
- The backdoor reportedly sends location and identification information to company servers without user consent.
- Alibaba will ban 'Cloud Code' from July 10 citing security concerns.
- Anthropic engineer Thariq Shihipar stated the data tracking was part of a test to prevent misuse and will be removed in the next version.
