GitHub's New Data Collection Policy for Copilot Users

GitHub has announced a significant policy change regarding the use of interaction data from its Copilot users. Starting April 24, 2026, all personal account users of GitHub Copilot—specifically those with Copilot Free, Copilot Pro, and Copilot Pro+ accounts—will be automatically enrolled in a data collection program aimed at training AI models. Users can opt out of this data collection through their account settings, but the default enrollment raises concerns about user consent and data privacy.

GitHub Enables Copilot Data Collection for AI Training by Default With Opt-Out SettingThe Register UK

The data collected by GitHub includes various forms of user interactions, such as input and output data, code snippets, comments, documentation, file names, and repository structures. This information is intended to enhance the performance of GitHub's AI models, improving code suggestions and bug detection capabilities. However, the announcement does not clarify the minimum interaction threshold for data collection or how the data will be anonymized before use. Additionally, users on Copilot Business and Copilot Enterprise plans are exempt from this default data collection, but details regarding their data handling remain unspecified.

GitHub Enables Copilot Data Collection for AI Training by Default With Opt-Out SettingThe Register UK

Mario Rodriguez, GitHub's Chief Product Officer, emphasized the benefits of data collection, stating that user participation would help models better understand development workflows and improve the accuracy of code suggestions. He noted that the initial Copilot models were built using publicly available data and that incorporating interaction data from Microsoft employees had already led to performance improvements. GitHub claims that this practice aligns with "established industry practices," particularly in the U.S., where opt-out policies are more common than the opt-in requirements seen in Europe.

The Register UKGitHub Enables Copilot Data Collection for AI Training by Default With Opt-Out Setting

The response from the GitHub community has been largely negative, with users expressing concerns over the implications of this policy change. Feedback on the announcement included 59 thumbs-down votes compared to just three positive reactions. Critics argue that the automatic enrollment undermines the concept of privacy associated with GitHub's "private" repositories, as code snippets from these repositories could potentially be collected for training purposes. The lack of clarity regarding data anonymization and the absence of a clear timeline for when data collection began further exacerbate user apprehensions.

The Register UK

While GitHub's announcement outlines the data collection practices, it lacks specific details on how sensitive or proprietary information will be protected. There is also no mention of whether interactions prior to the announcement will be included in the data collection. This ambiguity leaves users uncertain about the extent of data usage and the safeguards in place.

GitHub Enables Copilot Data Collection for AI Training by Default With Opt-Out Setting

As GitHub moves forward with this policy change, users will need to navigate their account settings to opt out if they choose. The company may face ongoing scrutiny and potential backlash from its user base, which could influence future adjustments to its data collection practices.

The Register UK