Anthropic to Use Claude Chatbot Conversations as Training Data

Starting October 2025, Anthropic will use user conversations with its **Claude chatbot** for training models, unless users opt out. This policy shift updates prior restrictions. New users are asked to consent at sign-up; existing users will see toggles in settings. Claude’s past policy avoided using user chats for training. The move aims to improve accuracy, but raises concerns about data privacy and user transparency.