GitHub’s June 1, 2026 update puts GitHub Copilot usage-based billing into effect across Copilot plans. In the same rollout, GitHub confirmed that Copilot code review now consumes GitHub Actions minutes in addition to GitHub AI Credits, and introduced broader budget controls for organizations and enterprises.

That is not just a pricing footnote. For QA engineers and automation teams using Copilot for pull request review, test maintenance, and debugging, the cost model now tracks closer to actual agent usage. Teams that review large private repositories or encourage AI-assisted PR checks need to revisit both spend controls and how they measure value.

What changed on June 1

  • Usage-based billing is live: GitHub says all Copilot plans now bill based on GitHub AI Credits consumed.
  • Private repo code review costs changed: Copilot code review now uses GitHub Actions minutes as well as AI Credits.
  • Budget controls expanded: GitHub says user-level budgets are generally available for organizations and enterprises.
  • Runner controls improved: organization admins can set a default runner for Copilot code review instead of configuring repositories one by one.

GitHub’s earlier April 27 notice gave teams advance warning that the code review agent runs on GitHub Actions and that private-repository reviews would begin consuming Actions minutes on June 1. The June 1 changelog confirms that switch is now active.

Why GitHub Copilot usage-based billing matters for QA engineers

Many QA teams use Copilot beyond code completion. It may review test-related pull requests, explain failures, suggest locator fixes, or help refactor flaky automation. Under the new model, those activities can affect two different meters when they involve private repositories and code review automation.

  • PR review becomes easier to misprice: a team may see useful review comments, but the real cost now includes both Copilot usage and Actions runtime.
  • Large automation repos need tighter guardrails: suites with many pull requests, generated snapshots, or heavy test fixtures can drive more review activity than teams expect.
  • Admin defaults matter more: runner choices and budget settings are now part of test-tool governance, not just platform administration.

One important measurement gap to know

GitHub’s Copilot usage metrics documentation says the metrics are derived from IDE telemetry and do not include Copilot code review or Copilot CLI activity. For QA leads, that means a dashboard showing healthy adoption trends is not the full billing picture if engineers heavily use PR review agents or terminal workflows.

In practice, teams should read Copilot usage dashboards together with billing data and GitHub Actions usage, especially if AI-assisted review is part of the release process.

What QA teams should check this week

  • Audit private-repo review volume: identify which repositories use Copilot code review and estimate how often reviews run on active test automation projects.
  • Set budget thresholds before scale-up: confirm enterprise, organization, or user-level budgets are aligned with expected AI-assisted review usage.
  • Review runner policy: if your team has special security or performance constraints, verify whether the default runner for Copilot code review matches them.
  • Separate quality from convenience: measure whether Copilot reviews are catching meaningful test risks such as missing assertions, brittle waits, or regression gaps before expanding usage.

Why this matters for QA engineers

This update matters because it changes how AI-assisted testing work is governed in real teams. A tester asking Copilot for code review on a private repo is no longer just consuming an abstract assistant allowance. They may also be using billable GitHub Actions capacity. QA leaders who treat Copilot as part of the automation toolchain now need budget visibility, runner decisions, and review policies alongside prompt quality.

Sources