GitHub Will Use User Data to Train Microsoft's AI

GitHub has confirmed that user data from Free, Pro and Pro+ accounts may be used to train Microsoft’s AI models, but stressed that account holders can opt out if they prefer. The move expands the dataset available to Microsoft’s AI efforts and has stirred conversation in developer communities about consent and code ownership.

According to GitHub’s latest notice, code, issues, pull requests and other repository activity could be included in model training unless an account explicitly disables that option. The company frames the change as part of improving automated coding tools and AI features that rely on large, diverse codebases to learn patterns and offer useful suggestions.

For many developers, the ability to opt out will be the most important detail. GitHub says opting out won’t break repositories or remove existing features, but it may limit how much tailored help users receive from AI-driven tools. Those who rely on Copilot-like suggestions might notice differences if their content is excluded from training pools.

Privacy advocates and some open source contributors have voiced concerns about the policy, especially around code that is public but created by individuals who may not want it used for commercial model training. GitHub emphasizes that the change follows existing licensing and terms, and that repository visibility and license choice still matter for how code can be reused.

If you’re a GitHub user wondering what to do: check your account’s data usage settings and review any project licenses you maintain. Opting out is an option for those who want to keep their work from feeding large models, while others may leave participation enabled to support broader AI improvements.

Overall, this update is another turn in the ongoing conversation about how platform data fuels AI development, and how companies balance innovation with user control.