Exercise - Reconcile Copilot and agent costs across surfaces

  • 10 minutes

Leadership's readout needs a reconciled cost picture: Copilot's metered spend, and the Woodgrove fraud-investigation agent's separate Azure spend.

Important

Prerequisites for this exercise:

  • A Microsoft 365 tenant with AI Administrator or License Administrator access to Copilot billing policies.
  • An Azure subscription with Cost Management read access to a lab Foundry agent's resource group (for example, the one backing the Woodgrove fraud-investigation agent).

Task 1: Review Copilot's pay-as-you-go spend in the admin center

  1. Sign in to the Microsoft 365 admin center at admin.microsoft.com, then go to Copilot > Settings > Pay-as-you-go (or search for pay-as-you-go if the entry point differs in your tenant).
  2. Open the billing policy covering Copilot metered usage, and review its Budget and spending view.
  3. Note the current spend against the configured budget, and the metered service it's tracking.

Task 2: Review the agent's Azure Foundry spend

  1. Sign in to the Azure portal at portal.azure.com, and open the resource group backing the Woodgrove fraud-investigation agent.
  2. Select Cost Management > Cost analysis, and review the resource group's spend for the current billing period.
  3. Compare this to the Copilot pay-as-you-go spend from Task 1. Reflect: these are two genuinely separate cost surfaces—Copilot's metered spend lives in the Microsoft 365 admin center's billing policy, while a Foundry agent's compute and model costs live in Microsoft Cost Management against its own resource group. Reconciling "total AI spend" for leadership means checking both, not assuming one number covers everything.

Task 3: Configure a budget alert on the Copilot billing policy

  1. Back in the Microsoft 365 admin center's billing policy for Copilot, select Budget > Add budget (or Edit budget if one already exists).
  2. Set a budget amount appropriate for the current spend trend, and add an alert threshold (for example, 80%).
  3. Add a recipient group—an email distribution list or a set of admin accounts—to receive the alert notification.
  4. Save the budget. Reflect: this is the same discipline as the Azure budget alert you may have configured for AI services in an earlier module—Copilot's pay-as-you-go spend needs the same proactive guardrail as any other metered Azure-backed AI service.

You've now reviewed Copilot's metered spend in the admin center, reviewed the Woodgrove agent's separate Azure Foundry costs, and configured a budget alert so neither cost surface goes unmonitored between now and the next leadership readout.