Anthropic Removes Hidden User Code in Claude Code, Drawing Attention
A hidden code in Claude Code ran silently from April 2nd to July 1st without any update logs or official documentation, and most users were unaware. The code was discovered by Reddit users who decompiled the binary files, and security researchers conducted reverse engineering analysis on multiple versions. Overseas media published an in-depth report on July 1st. Its core function is to specifically identify domestic users or those accessing through proxy interfaces of Chinese AI enterprises, and the entire mechanism is suspected to be used for targeted monitoring and regulation.