GitHub Copilot Personal Plans to Use User Code for AI Training by Default Starting April 24; Opt Out Manually

GitHub recently updated Copilot's data usage policy. Starting April 24, interaction data from Copilot Free, Pro, and Pro+ personal users will be used for AI model training and improvement by default, unless users manually opt out in settings. Copilot Business and Enterprise users are not affected, and users who previously opted out do not need to reapply their preferences.

Collected data includes code outputs accepted or modified by users, input snippets, code context, comments and documents, repository structure, and interaction records. When using Copilot, real-time code interaction data from private repositories may also be used for training. Data is only shared with affiliated companies like Microsoft and not with third-party AI vendors. GitHub states this move aims to improve model performance. Users can turn off data sharing via GitHub's privacy settings.