GitHub Copilot gets more ways to request code review

GitHub has updated Copilot code review, the feature where the AI assistant Copilot automatically reviews code in pull requests and suggests fixes, just as a colleague would during a code review.

Previously, the personal settings for automatic code review were only available to users on the more expensive Copilot plans Pro, Pro+ and Max. Now a unified settings page has been rolled out to all Copilot plans, including Business and Enterprise, which many companies use.

On the new page, you can turn on automatic code review so that Copilot starts on its own when you create a pull request (a request to have your code merged into a project). You can also choose whether Copilot should review code on new pushes (when you upload changes) or on draft pull requests.

In addition, users can choose how thoroughly Copilot should work: "Lite" for a faster, more superficial review, or "Balanced" for a more balanced and thorough analysis. This applies both when Copilot starts the review automatically and when you request it manually.

For companies, the update means IT administrators can now set a common standard for how thoroughly Copilot should review code across the entire organization. Individual teams and projects can still choose to configure it differently, however, if the common standard does not fit.

For ordinary developers, it mainly means more freedom to adjust when and how thoroughly the AI looks through their code, without requiring a more expensive subscription.

What this means for you

If you use GitHub Copilot to write code, you can now decide for yourself when and how thoroughly the AI should review your pull requests, regardless of which plan you have. Go to the new settings page, turn on automatic code review and choose between Lite or Balanced as needed. If you are an IT administrator, you can set a common standard for the entire organization, while teams can still adjust it themselves.