# AI and Power BI.

URL: https://levystreet.com/integrations/power-bi/
Updated: 2026-10-19

An AI Power BI integration guide for NZ: Copilot capacity rules, executeQueries, Power BI MCP servers and Fabric IQ, and narratives with reconciled metrics.

## Short answer

An AI Power BI integration uses a model to explain, query or document your semantic models, through Copilot in Power BI, DAX queries over the REST API, or Microsoft's Power BI MCP servers. For New Zealand organisations the safe pattern for management reporting is to compute and reconcile every number in DAX and code first, then let the model write commentary that a finance lead approves.

- Copilot in Power BI needs Fabric F2 or Premium P1 capacity; Pro licences alone are not enough.
- Standalone Copilot and Copilot in apps are not yet available in the New Zealand North region.
- Microsoft processes Copilot for Australian capacities in the US, after cross-geo processing is enabled.
- executeQueries returns at most 100,000 rows and can report success when results are cut short.
- Microsoft points natural-language Q&A to Fabric IQ, which respects row-level security.

## What Copilot in Power BI does (October 2026)

Copilot in Power BI summarises reports, answers questions about data, drafts report pages and DAX, writes narrative visuals and adds summaries to subscriptions. Per [Microsoft's overview](https://learn.microsoft.com/en-us/power-bi/create-reports/copilot-introduction), last updated 25 August 2026, the report Copilot pane is generally available while the standalone Copilot and Copilot in apps are in preview. It needs paid Fabric capacity (F2 or higher) or Power BI Premium (P1 or higher); a Pro or Premium Per User licence alone is not enough, and trial capacities are excluded. Microsoft also warns that without preparing the semantic model for AI, Copilot can return generic or inaccurate results.

Two New Zealand points matter. The standalone Copilot and Copilot in apps are not yet available in the New Zealand North region. And for capacities outside the US and EU boundary, Copilot is off by default until a Fabric admin allows cross-geo processing; [Microsoft's table](https://learn.microsoft.com/en-us/fabric/fundamentals/copilot-fabric-overview) shows Australian capacities processed in the US. NZ list prices on 19 October 2026, excluding GST and paid yearly: Pro NZ$22.60 and Premium Per User NZ$38.80 per user per month.

## Integration routes and when to use each

| Route | What it gives you | Use it when |
|---|---|---|
| Copilot in Power BI | Summaries, Q&A, report and DAX drafting | You have F2 or P1 capacity and have decided on cross-geo processing |
| REST API `executeQueries` | DAX queries against a semantic model, returned as JSON | Your own pipeline needs certified measures |
| Power BI Authoring MCP server | Create, change and document model objects; hosted preview or local (not on macOS) | Model developers refactoring, documenting or applying best practice |
| Fabric IQ MCP | Natural-language answers grounded in semantic models, respecting row- and object-level security | Grounding Copilot Studio, Microsoft Copilot or custom agents in governed data |
| Power BI Consumption MCP | An earlier query endpoint kept for existing integrations | Avoid for new builds |

## Automation 1: a monthly narrative with reconciled metrics


- **Inputs**: About a dozen certified measures pulled with `executeQueries` for this month, last month and the same month last year.
- **Reconciliation in code**: Revenue and expenses must tie to the ledger trial balance, and debtors to the aged receivables report, within set tolerances. A failed check stops the pack and shows the variance.
- **Model step**: Commentary written only from the numbers table plus notes from budget holders. Each sentence carries the metric ID it relies on.
- **Validator**: Every figure in the text must match the table after rounding; percentage changes are recomputed in code.
- **Human approval**: The finance lead edits and approves; the CFO signs off.
- **Output**: The approved text is stored with its data snapshot and version, then published to the report or the board pack. Nothing writes to the semantic model.
- **Failure and recovery**: Fix the source, rerun, and keep the last approved version until a new one is signed off.


| Metric ID | DAX measure | Reconciles to | Tolerance |
|---|---|---|---|
| REV | [Revenue] | Ledger revenue accounts | NZ$1 |
| GM | [Gross Margin %] | Recomputed from revenue and cost of sales | 0.1 points |
| DR60 | [Debtors 60+ days] | Aged receivables report | NZ$1 |

Many NZ organisations run a 1 April to 31 March financial year, so check that year-to-date measures use your fiscal calendar. The workflow is mapped in [management reporting commentary](/industries/financial-services/management-reporting-commentary/), and our [AI reporting](/services/ai-reporting/) service builds it end to end.

## Automation 2: documenting a model with the Authoring MCP server


1. **Connect safely.** Point the [Authoring MCP server](https://learn.microsoft.com/en-us/power-bi/developer/mcp/power-bi-authoring-mcp) at a copy of the model or at Power BI Project files under source control.
2. **Inventory.** Ask for every measure with its dependencies, and flag duplicates, unused measures and missing descriptions.
3. **Review.** A developer accepts proposed descriptions and fixes as a pull request. Nothing is applied to production directly.
4. **Benefit twice.** Clear descriptions also help Copilot, which Microsoft says depends on a model prepared for AI.


## Permissions and approval design

For [executeQueries](https://learn.microsoft.com/en-us/rest/api/power-bi/datasets/execute-queries), a tenant admin must enable "Dataset Execute Queries REST API", the caller needs read and build permission on the dataset, and the token needs Dataset.Read.All. Service principals need their own tenant setting and are not supported on datasets with row-level security or SSO, so a reporting bot on a secured model must run as a named user or use an unsecured certified model built for the pack. The hosted Authoring MCP server needs the admin setting "Users can use the Power BI Model Context Protocol server endpoint".

Microsoft says REST API request and response data may be processed outside your tenant's home region. If the model contains payroll or customer data, record that in your privacy impact assessment; see [data residency for AI systems](/engineering/data-residency-for-ai-systems/).

## Costs and limits (as listed 19 October 2026)

- **executeQueries**: one query per call, one table per query, at most 100,000 rows or 1,000,000 values and 15 MB, and 120 queries a minute per user.
- **Copilot**: needs capacity on top of user licences; new capacity can take up to 24 hours to be recognised.
- **Model cost, worked example**: a monthly pack at 20,000 input and 2,000 output tokens on Claude Sonnet 5.5 costs about US$0.06 per draft. The real cost is modelling and reconciliation work.

## Where it goes wrong

- **Silent truncation.** A query that exceeds the row limit returns HTTP 200 with partial data and an error in the body. Always check the error field.
- **Invented figures.** Any number the model computes itself will eventually be wrong. Validate every figure.
- **Calendar mismatch.** Year-to-date on a calendar year looks plausible and is wrong for a March balance date.
- **Unprepared models.** Cryptic measure names give Copilot and custom agents poor answers.

Most packs pull from the ledger too; see our [Xero](/integrations/xero/) and [Microsoft 365](/integrations/microsoft-365/) guides.

## Questions people ask

### Do we need Fabric capacity to use Copilot in Power BI?

Yes. Microsoft's requirements, last updated 25 August 2026, are a paid Fabric capacity of F2 or higher or Power BI Premium P1 or higher. A Pro or Premium Per User licence is not enough on its own, and trial capacities are not supported. An admin setting must also allow Copilot, and new or resized capacity can take up to 24 hours to be recognised.

### Is Power BI Copilot available in New Zealand?

Partly. Microsoft lists New Zealand North among regions where the standalone Copilot and Copilot in apps are not yet available, so check its supported-region list for your capacity. For capacities outside the US and EU boundary, Copilot is off by default until a Fabric admin allows data to be processed outside the region, and Microsoft's table shows Australian capacities processed in the US.

### Can ChatGPT or Claude query our Power BI data?

Microsoft now points natural-language questions over Power BI data to Fabric IQ, which has an MCP endpoint and respects the row-level and object-level security on your semantic models. The Power BI Authoring MCP server is for building and documenting models, not answering business questions. Check your tenant settings and licensing before connecting any external assistant to finance data.

### Can AI write our monthly management report commentary?

It can draft it, provided the numbers are not its job. Pull certified measures with DAX, reconcile them to the ledger in code, and give the model only that table plus notes from budget holders. Check every figure in the draft against the table automatically, and have the finance lead approve the text. Explanations of why something moved must come from people, not the model.


## Sources

- [Copilot for Power BI overview](https://learn.microsoft.com/en-us/power-bi/create-reports/copilot-introduction)
- [Overview of Copilot in Fabric](https://learn.microsoft.com/en-us/fabric/fundamentals/copilot-fabric-overview)
- [Power BI pricing (NZ)](https://www.microsoft.com/en-nz/power-platform/products/power-bi/pricing)
- [Power BI REST APIs](https://learn.microsoft.com/en-us/rest/api/power-bi/)
- [Datasets: Execute Queries](https://learn.microsoft.com/en-us/rest/api/power-bi/datasets/execute-queries)
- [Power BI MCP servers overview](https://learn.microsoft.com/en-us/power-bi/developer/mcp/mcp-servers-overview)
- [Power BI Authoring MCP server](https://learn.microsoft.com/en-us/power-bi/developer/mcp/power-bi-authoring-mcp)
- [Pricing](https://platform.claude.com/docs/en/about-claude/pricing)

Published by Levy Street (https://levystreet.com), Wellington, New Zealand.