Alibaba bans Claude Code after alleged China-detection code surfaces

Alibaba has banned employees from using Anthropic’s Claude Code for work, effective July 10, after security researchers alleged the coding agent included hidden logic to identify users in China or connected to Chinese AI labs. The company reportedly classified Claude Code as high-risk software and told staff to move to Qoder, its in-house AI coding platform, with Chinese outlets saying the directive also covers other Anthropic products including Sonnet, Opus, and Fable.

The dispute centers on a June 30 Reddit post claiming reverse-engineered Claude Code contained obfuscated detection logic since version 2.1.91, released on April 2. The code allegedly checked proxy use, time zones such as Asia/Shanghai and Asia/Urumqi, and a hardcoded list of Chinese domains and AI lab identifiers, then encoded findings through subtle changes in the system prompt sent to Anthropic’s servers.

Anthropic has not issued a formal statement, but Claude Code engineer Thariq Shihipar described the mechanism as “an experiment we launched in March” to curb unauthorized resellers and model distillation. The clash follows Anthropic’s June 10 letter to U.S. Senate Banking Committee leaders accusing operators affiliated with Alibaba’s Qwen lab of using nearly 25,000 fraudulent accounts to generate 28.8 million exchanges with Claude between April 22 and June 5, allegations Alibaba has denied.