Anthropic Removes Hidden Claude Code Tracker After Researchers Discover Privacy Concerns

According to Decrypt, Anthropic removed hidden tracking markers from Claude Code in June after developer Thereallo discovered the coding assistant was using undisclosed signals to identify users' location, proxy use, and possible links to Chinese AI labs. The tracking system, embedded in system prompts using Unicode markers and encoded domain lists, was designed to flag users attempting to bypass restrictions or extract model capabilities.

Anthropine engineer Thariq Shihipar stated the feature was introduced in March as an experiment to prevent account abuse and detect AI model distillation attacks. The company has since implemented stronger mitigations and rolled back the tracking system.