GitHub Adds AI Credit Pool Controls to the Billing UI for Copilot
GitHub now lets admins manage AI credit pools for cost centers directly in the billing UI, no REST API required. Available for Copilot Business and Enterprise.
GitHub has added AI credit pool management directly to its billing UI, available as of July 20, 2026. Previously, admins on Copilot Business and Copilot Enterprise (GitHub Enterprise Cloud) had to call the GitHub REST API to configure these pools. Now the same controls are accessible when creating or editing a cost center in the browser, including an overage policy that either blocks usage at the limit or permits additional spend.
What happened
| Detail | Fact |
|---|---|
| Feature available | July 20, 2026 |
| Plans supported | Copilot Business, Copilot Enterprise |
| Platform required | GitHub Enterprise Cloud |
| Previous method | GitHub REST API only |
| Pool limit set by | GitHub (calculated from assigned licenses automatically) |
GitHub shipped a billing UI update that brings AI credit pool management inside the cost center editor. Before this change, Copilot admins had to use the REST API to toggle a pool on or off and configure its behavior. Now the same toggle appears directly when you create or edit a cost center in the browser.
An AI credit pool limits how many included AI credits a cost center can consume to only what its own assigned Copilot licenses cover. This keeps individual teams or departments from drawing on credits funded by other groups’ licenses.
How the limit and overage policy work
GitHub calculates the pool limit automatically from the number of Copilot licenses assigned to that cost center. As you add or remove licenses, the limit adjusts to match. You never type a number yourself.
When a cost center reaches its pool limit, you choose one of two behaviors:
- Block further included usage. No additional AI credit consumption is allowed once the pool is exhausted.
- Allow overages. Usage continues but counts as additional spend, if your enterprise has overages enabled.
How does an AI credit pool differ from a cost center budget?
These are two separate controls, and GitHub lets you set both on the same cost center. The AI credit pool governs included credits tied to licenses. A cost center budget, by contrast, caps metered charges that accrue after the pool runs out. Think of the pool as the first guardrail and the budget as the second one.
Why it matters
For any organisation running GitHub Enterprise Cloud with multiple teams sharing a Copilot subscription, this closes a real management gap. Before, keeping teams within their own license allocation required REST API calls, which meant either scripting or relying on engineers who had API access. The billing UI change hands control to finance and operations staff who already live in that interface.
The automatic limit calculation also removes a common error vector. Manually syncing license counts to credit caps is error-prone, especially when teams fluctuate. GitHub now owns that arithmetic.
For businesses exploring how to govern AI spend across departments, this kind of per-team guardrail is increasingly standard. If you are evaluating broader AI integration strategies for your organisation, cost controls at the team level are worth building into your planning from the start.
Our take
This is a sensible, overdue quality-of-life fix. Requiring REST API access to manage spend controls was always an odd friction point for what is fundamentally an administrative task. Moving it into the billing UI is the right call.
The automatic limit calculation is the most practical part. Anyone who has tried to keep a manually entered credit cap in sync with a fluctuating headcount knows how quickly that gets stale. GitHub taking over that calculation removes a whole category of “why is the limit wrong” support tickets.
The dual-control design (pool plus budget on the same cost center) is also worth noting. It lets you build a two-layer spending policy without any custom tooling. That said, the overage option only works if your enterprise has overages enabled at all, so check that setting before assuming teams can flex past their pool.
For a broader look at what is changing across AI billing and governance tools, see our AI and web news coverage.
Frequently asked questions
What is an AI credit pool in GitHub Copilot?
An AI credit pool limits how many included AI credits a cost center can use to only what its own assigned Copilot licenses cover. It stops one team from consuming credits funded by another team's licenses.
How do I set the AI credit pool limit in GitHub?
You don't set a number manually. GitHub calculates the pool limit automatically based on the number of Copilot licenses assigned to a cost center and adjusts it whenever you add or remove licenses.
What happens when a cost center reaches its AI credit pool limit?
You can choose to either block further included usage or allow it to continue as additional spend, provided your enterprise has overages enabled.
What is the difference between an AI credit pool and a cost center budget on GitHub?
An AI credit pool caps the included credits tied to a cost center's Copilot licenses. A cost center budget caps metered charges that accrue after the pool is exhausted. You can set both on the same cost center.
