How to Set Up Portal AI in PowerBI Portal: Step-by-Step Guide (2026)

Portal AI is included in PowerBI Portal, there is no separate installation. What you need to do is enable it at the environment level, then activate it report by report.

This guide follows the official setup flow. If you have not yet read the overview of what Portal AI is and what it does, the article 'What Is Portal AI? How to Ask Questions About Your Power BI Data in Plain Language' covers the fundamentals first.

Setup Overview

The full setup involves four steps:

Before You Start

• You have the Master role in PowerBI Portal.
• You have access to the Power BI admin portal to configure permissions.
• At least one Power BI report is already configured in PowerBI Portal.

Step 1: Enable Portal AI at Environment Level

This is a one-time setup done by a Master. It activates Portal AI for the entire PowerBI Portal environment and grants the necessary permissions in Power BI.

There are two parts to this step: enabling Portal AI in your environment settings, and enabling the required permissions in the Power BI admin portal. Both are needed before Portal AI will work on any report.

For the full permission requirements and the exact settings to enable in the Power BI admin portal, refer to the official documentation:

📄 docs.powerbiportal.com/portal-ai/enabling-portal-ai/

Step 2: Configure Portal AI on Each Report

Portal AI is activated per report, not globally. This means you choose exactly which reports offer the AI question feature to users.

To activate Portal AI on a report:

• Open the report in the PowerBI Portal back-office.
• Go to the Portal AI tab in the report settings.
• Toggle Enable Portal AI on.
• In the Semantic Model Schema section, check that the Semantic Model ID is correct.
• Click Get/Update Schema to retrieve or refresh the semantic model schema.
• Save the report.

Portal AI uses the semantic model schema to understand the data structure behind the report. Without the schema, it cannot generate accurate answers. If the underlying dataset changes, return to this step and click Get/Update Schema again to keep it current.

Step 3: Manage Access (Optional)

By default, any user with access to the report can use Portal AI on it. If you want to restrict Portal AI to specific users or groups, you can configure this in the access settings for each report.

This step is optional. If you skip it, Portal AI is available to everyone who can view the report.

For the full details on how to configure per-user and per-group access, refer to:

📄 docs.powerbiportal.com/portal-ai/managing-ai-access/

Testing Before You Open It Up

Before Portal AI reaches end-users, test it yourself from the front-office.

Log into the portal as a regular user and navigate to a report where Portal AI is enabled. You should see a chat button in the bottom-right corner of the report page. Type a straightforward question based on the report data and verify that:

• The answer returns within a few seconds
• The answer reflects data consistent with what you see in the report
• If the report uses RLS, the answer only includes data within your user's access scope

If the chat button does not appear

• Verify that Portal AI is enabled at environment level (Step 1) and that the Power BI admin portal permissions are configured.
• Check that the report has Portal AI toggled on with a schema loaded (Step 2).
• Check whether the user has been granted access to Portal AI on this report (Step 3).

Note: reports that use Row-Level Security require a Premium or Embedded capacity to work with Portal AI. Without capacity, Portal AI will show an error for RLS-enabled reports.

Frequently Asked Questions

What languages does Portal AI support?

Portal AI answers in the same language you use in the conversation. Response quality may vary depending on the language, with the best results in widely spoken languages.

Does enabling Portal AI cost extra?

Portal AI usage is covered by credits. All plans include a 31-day trial with 100 credits. Usage beyond the trial is included in the Enterprise plan. Usage history is available in License & Billing.

What happens if the semantic model schema is out of date?

Portal AI may give inaccurate or incomplete answers. Return to the Portal AI tab of the report in the back-office and click Get/Update Schema to refresh it after any significant changes to the underlying dataset.

Schema discovery failed. What should I do?

The most common causes are permissions not yet configured in the Power BI admin portal, or settings that have not propagated yet. After making changes in the Power BI admin portal, wait up to 15 minutes and try again. If the issue persists, the Portal AI FAQ in the documentation covers additional troubleshooting steps.

Does Portal AI work with RLS-enabled reports?

Yes. Portal AI respects the RLS rules defined for each report in PowerBI Portal. Users only receive answers based on the same data they can see when viewing the report normally. Note that RLS with Portal AI requires a Premium or Embedded capacity without one, Portal AI will display an error on RLS-enabled reports.

Portal AI is included in PowerBI Portal. Start a free trial here.