GitHub has reversed its decision and will now train its AI models using customer interaction data, including inputs, outputs, code snippets, and context, starting April 24. This revised policy affects Copilot Free, Pro, and Pro+ users, while Copilot Business, Enterprise, students, and teachers are exempt. Users can opt out through their settings, though GitHub's Chief Product Officer, Mario Rodriguez, encourages participation for improved AI accuracy and bug detection. The company cites similar opt-out policies from Microsoft, Anthropic, and JetBrains, and claims that incorporating Microsoft employee data has already led to significant improvements in AI suggestion acceptance rates. This change redefines the privacy of private repositories, as code snippets can be collected for training if users enable the setting. Community reaction, based on emoji votes and comments, indicates widespread user disapproval.