Skip to main content

What are Power BI’s Copilot features?

An overview of the Copilot features available in Power BI today, what each one does, and where their practical limits are.

George avatar
Written by George
Updated over a month ago

💁 About this article
This article is part of Fugo’s Power BI knowledge base: a collection of resources answering common Power BI questions. We include notes throughout where Fugo’s integration may be helpful for displaying dashboards on digital signage.

Table of contents


Short answer

Power BI Copilot is an AI-assisted feature set inside Microsoft Fabric that helps users create report pages, generate narrative summaries, and query data using natural language. It works best when your data model is well-defined and your use case is exploratory rather than precise.

Copilot is available only on paid Fabric or Premium capacity (F64 / P1 and above) and must be enabled by a Fabric admin.


Report page generation from prompts

Copilot can generate entire report pages based on a natural language description of what you want to analyze.

Screenshot showing Copilot sample prompt for Customer segmentation.

Typical uses:

  • Creating a first draft of a report page

  • Exploring unfamiliar datasets

  • Rapid prototyping during early analysis

What to expect:

  • Reasonable visual selection

  • Basic field mapping

  • Limited ability to refine or “fix” what it creates

Copilot cannot reliably:

  • Modify complex existing visuals

  • Apply precise formatting rules

  • Replace deliberate data modeling or measure design


Editing existing report pages

You can ask Copilot to:

  • Add visuals to a page

  • Remove visuals

  • Swap visual types

  • Change fields used in a visual

This works best for structural changes, not visual polish. Styling, conditional formatting, and layout consistency are still largely manual tasks.

Screenshot of Copilot pane displaying a generated Power BI report page with visualizations.

Narrative summaries

Copilot can generate written summaries of:

  • A full report

  • A specific page

  • Individual visuals

This is one of the more consistently useful features, especially for:

  • Executive summaries

  • Read-only dashboards

  • Contextual explanations in shared environments

Screenshot showing select an option in the Create a narrative with Copilot dialog.

Natural language Q&A (via Copilot and Q&A visuals)

Copilot improves Power BI’s natural language querying by:

  • Suggesting synonyms for model fields

  • Helping users ask questions in plain language

  • Returning visuals or summaries based on those questions

Accuracy here depends heavily on:

  • Model design

  • Clear field naming

  • Carefully reviewed synonyms

Copilot does not understand your data automatically - it relies on metadata you provide.

Screenshot that shows Copilot making synonym suggestions.

Most important requirements and limitations

Copilot in Power BI requires:

  • Fabric capacity F64 or higher (or Power BI Premium P1+)

  • Tenant-level admin enablement

  • Paid capacity (trial SKUs are not supported)

Current limitations include:

  • No support for custom visuals

  • No reliable styling or formatting control

  • Limited DAX and model restructuring capability

  • Inconsistent results with complex logic


Where Copilot fits and where it doesn’t

Copilot works best as:

  • A starting point for analysis

  • A helper for narrative explanations

  • A convenience layer for non-technical users

It is not a replacement for:

  • DAX expertise

  • Data modeling discipline

  • Purpose-built analytics delivery (dashboards, displays, apps)


Power BI on digital signage

Copilot can help create and summarize reports, but it doesn’t handle how those reports are displayed in shared environments.

For teams using Power BI on office TVs, operations floors, or public-facing screens, Fugo’s Power BI app provides a reliable delivery layer: unattended playback, refresh control, and screen-native layouts built for digital signage rather than interactive analysis.

Did this answer your question?