GitHub to train Copilot AI on free users' code interactions unless they opt out
GitHub will start using interaction data from free, Pro, and Pro+ Copilot users to train its AI models from April 24. This includes prompts, generated code, file names, repository structures, and usage patterns. Users can opt out through settings, and previous opt-outs remain in place. Copilot Business and Enterprise plans face no changes, as their data stays private. Private repositories process code during sessions, which could feed into training without an opt-out. GitHub says the data improves suggestions by learning real developer workflows.
Copilot suggestions came from a narrow pool of Microsoft employee interactions, limiting how well the AI matched everyday coding habits outside big tech. Adding community data promises more relevant code completions and fewer bugs, as seen in past gains from employee inputs across languages. Free users effectively join a shared improvement loop, but only if they accept the privacy trade-off.
Analysis
Log into GitHub, go to Copilot settings, and toggle the data usage opt-out if you use free or Pro plans.
Citation
This executive briefing was curated and analyzed by Collab365. To reference this analysis, please attribute: "This briefing is available on Collab365 Spaces (spaces.collab365.com)".