JetBrains Air adds GitHub Copilot for local and cloud coding tasks

JetBrains Air added GitHub Copilot for local and cloud coding workflows.

**Dek:** The November 3 integration uses Agent Client Protocol and an existing Copilot subscription. Cloud tasks also require Air Teams access and an administrator’s approval.

GitHub added Copilot to JetBrains Air’s agent picker on November 3. A developer can connect an existing Copilot account in the Air plugin and delegate tasks from a JetBrains IDE. The integration uses Agent Client Protocol, which gives the agent a way to work through Air’s interface. GitHub also says developers can keep using the existing GitHub Copilot plugin alongside the Air integration and choose which workflow suits a task. [GitHub announcement](https://github.blog/changelog/2026-11-03-github-copilot-is-now-available-in-jetbrains-air/)

Local and cloud tasks have different prerequisites

For a local task, the developer installs Air in the IDE, selects GitHub Copilot and connects the account. To use cloud execution, the developer also needs a JetBrains account in an Air Teams-enabled business organization and Copilot must be enabled by the Air administrator. In the browser, the setup includes connecting GitHub and choosing a repository and remote branch. The IDE plugin offers Cloud in its environment selector. GitHub describes these as routes into the same Copilot cloud integration. [GitHub announcement](https://github.blog/changelog/2026-11-03-github-copilot-is-now-available-in-jetbrains-air/)

Cloud tasks run in a Docker container on a fresh virtual machine. GitHub says traffic goes through a proxy and state is not carried from task to task. A cloud task can install dependencies and run tests without changing the workstation. That also means a project relying on a developer’s machine-specific configuration may need a reproducible setup before the same task can work remotely. Cloud execution changes where the task runs. It does not answer every question about which repository data and credentials it can reach. [GitHub announcement](https://github.blog/changelog/2026-11-03-github-copilot-is-now-available-in-jetbrains-air/)

Usage oversight still spans both products

GitHub says Copilot cloud tasks use the Copilot plan rather than JetBrains AI credits. Air may discover a missing subscription or unavailable credits only when it starts the environment. The user cannot select a personal or organizational Copilot plan per task through Air. Air administrators can disable Copilot through policy, but Air’s analytics and spending controls do not cover Copilot usage. Teams should identify which account and reporting system they rely on before moving recurring workloads. [GitHub announcement](https://github.blog/changelog/2026-11-03-github-copilot-is-now-available-in-jetbrains-air/)

A useful trial is to give the same small repository task to the local and cloud paths, then compare setup, test results, permission prompts and how a user reviews the proposed edits. That can show whether Air improves the daily workflow before an administrator makes it available to a larger group.