Anthropic Claude user conversations exposed via public indexing

Hundreds of private user interactions with the Claude AI chatbot were found to be publicly accessible, likely due to improper link sharing or indexing configurations. It remains unclear if the exposure resulted from a platform-level vulnerability or user-side link management, but the incident highlights significant privacy risks in AI session handling.