- Updated: March 27, 2026
- 2 min read
GitHub Announces AI Model Training on Private Repositories – What It Means for Developers
GitHub has rolled out a new policy that will allow its AI systems, including Copilot, to train on code from private repositories unless each user explicitly opts out. The change applies to Free, Pro, and Pro+ accounts, while Business and Enterprise plans remain exempt. This move has sparked a heated debate across the developer community, raising concerns about code privacy, intellectual property, and security.
According to the announcement, GitHub will automatically collect usage data from private repos—code snippets, surrounding context, and metadata—to improve its AI models. Users can disable the feature through a per‑user toggle in their settings; however, there is no organization‑wide opt‑out option. The policy will take effect on April 24, and developers who do not opt out by that date will have their private code included in the training set.
Key points of the policy include:
- Automatic data collection from private repositories for Free, Pro, and Pro+ users.
- Business and Enterprise customers are excluded from the data collection.
- Opt‑out is available only at the individual user level, not at the organization level.
- Potential legal and security implications for proprietary code.
The developer community has responded with a mix of criticism and support. Critics argue that the policy could expose sensitive code to unintended parties and undermine trust in GitHub’s platform. Proponents claim that broader data access will make AI tools like Copilot more accurate and useful.
For a deeper dive into the technical and legal ramifications, read the original discussion on Hacker News. You can also explore related articles on our site, such as AI and Code Security and Latest GitHub Copilot Updates.
Stay informed about the evolving landscape of AI in software development by following our blog and subscribing to our newsletter.