Researcher 'Thereallo' finds hidden tracker in Anthropic's Claude Code AI
A security researcher known as "Thereallo" found hidden tracking code inside Anthropic's Claude Code AI.
The code quietly logged users' time zones and whether they were using proxy servers, supposedly to spot links to certain Chinese AI labs.
This discovery has sparked real concerns about how transparent and ethical Anthropic is with its tech.
Anthropic says tracking was March experiment
Anthropic says the tracking was just a March 2026 experiment to stop account abuse and model copying.
But critics are calling out the double standard, since Anthropic has accused Chinese firms of stealing its models, while it built its own tech using scraped online content without permission.
"Thereallo" warns that invisible trackers like this hurt user trust and says "the correct reaction is scrutiny."