Secret Claude tracker shocks users after Anthropic’s anti-surveillance stance

Users recently discovered a hidden monitoring feature in the Claude AI interface that logged their interactions, prompting accusations that Anthropic violated its own privacy-focused branding. An Anthropic engineer confirmed the tool was part of a limited internal experiment and has since been disabled, though the incident sparked broader concerns regarding data transparency. The situation highlights a growing tension between user expectations of privacy and the technical diagnostic practices employed by AI companies to refine their models.