Microsoft-owned GitHub has released a new code referencing tool for GitHub Copilot. The new feature will inform developers when AI-powered code suggestions are taken from public repositories.
GitHub first unveiled the Copilot tool back in June 2021. The AI programming assistant allows developers to get code suggestions with natural language commands. Copilot has been trained on billions of lines of code from public repositories and other sources. The tool has been found to reproduce long sections of licensed code without providing credits to its authors.
Microsoft, GitHub, and OpenAI have been sued for violating federal copyright law and state business laws. The developers argued that GitHub Copilot generates code suggestions that are identical to its training data.
How GitHub Copilot code referencing feature works?
With code referencing, GitHub Copilot checks code suggestions against all public repositories on the platform. The feature shows code matches with information about each repository in the editor’s sidebar. It’s up to the developers to either accept or reject the suggestion or ask GitHub Copilot to rewrite the code.
“Some want to learn from others’ work, others may want to take a dependency rather than introduce new app logic, and still others want to give or receive credit for similar work. Whatever the reason, it’s nice to know when similar code is out there,” GitHub explained in a blog post.
According to GitHub, the code matches would occur in less than 1 percent of GitHub Copilot’s suggestions. However, the code referencing tool should help to improve transparency and ease the concerns raised by developers regarding potential legal issues.
GitHub notes that code referencing support is currently available in private beta, and developers can sign up to join the waitlist on this page. Meanwhile, the GitHub team is actively looking for user feedback to improve the feature in future releases.