Skip to content

Power BI

Power BI row-level security in practice

One report for the whole company, yet everyone sees only their own data. Here is how row-level security works in Power BI, how to design it dynamically, and the details that are easy to miss at go-live.

Published updated November 22, 2026 4 min read

Illustration: a dashboard with fictional data

What row-level security is for

Picture a sales dashboard. The sales director should see the whole business, regional managers only their region, and account managers only their own customers. Without proper security you end up maintaining several copies of the same report — and sooner or later they drift apart.

Row-level security (RLS) solves this with one data model and one report. Filters are applied based on who is signed in, so users never see other people's rows — not in visuals, not in tables and not in exports to Excel.

Static roles: the quick start

The simplest approach is static roles. In Power BI Desktop, under Manage roles, you define a role such as "Region North" with a filter like [Region] = "North" on the region table. After publishing, you assign users or Entra ID security groups to the role.

Static roles are easy to understand and test. Their weakness is maintenance: every new region or branch means a new role that someone has to look after. Fine for three regions; impractical for fifty account managers.

Dynamic RLS: one rule for everyone

Dynamic RLS uses the USERPRINCIPALNAME() function, which returns the signed-in user's login (typically their email address). You add a security table to the model, for example:

Email Region
anna@contoso.com North
ben@contoso.com South

Relate it to the region dimension and create a single role filtered by [Email] = USERPRINCIPALNAME(). Anyone listed sees their regions; anyone missing sees nothing. Permission management moves from role administration into data, and the security table can be fed from your HR system or CRM.

Tip: one user can have several rows, so a manager responsible for two regions needs no special role.

Hierarchies and exceptions

Requirements such as "managers see their own team and everyone reporting to them" appear quickly. DAX parent-child functions (PATH, PATHCONTAINS) can handle organisational hierarchies effectively, but they add complexity — design the hierarchy up front and test it against real data.

For executives who need to see everything, a separate unfiltered role is cleaner than exceptions buried in DAX.

Commonly missed details

  • Workspace roles. RLS applies to users with the Viewer role. Workspace Admins, Members and Contributors see all data. Anyone who should only read reports must not have a higher role.
  • Shared semantic models. RLS lives in the model, not the report. When other teams build reports on your model, security applies to them too — a strength, as long as everyone knows it.
  • Testing. Both Desktop and the Power BI service offer "View as" to test a role or a specific user. Include a user who is missing from the security table.
  • Performance. Complex filters on large tables slow down every query. Filter the smallest possible dimension and let relationships propagate the filter.
  • Hiding columns is not security. If some users must not see entire columns, such as margin, use object-level security (OLS) rather than simply hiding fields in a report.

A recommended approach

  1. Write a "who sees what" matrix in business language, not DAX.
  2. Decide whether static roles suffice or a security table is needed.
  3. Define where the security table comes from and who owns it.
  4. Assign workspace roles following the principle of least privilege.
  5. Test every role and edge case before rolling the report out.

Bottom line

Row-level security gives you a single source of truth while protecting sensitive data. Static roles work for a handful of groups; dynamic RLS scales to entire organisations. The most common mistakes are not in DAX but in workspace permissions and insufficient testing.

If you are working on report security or data model design, take a look at our Power BI implementation approach — security and roles are part of it from day one.

Choosing a BI tool?

Download the BI tool selection guide with a comparison and a checklist.

Illustration: a dashboard with fictional data

Power BI

Power BI vs. Tableau: choosing the right BI platform

Both tools produce great-looking dashboards. The real differences lie in total cost, how well each fits your existing stack, and who will use it every day. Here is a practical framework for the decision.

· 4 min read

Illustration: a dashboard with fictional data

Reporting

What does a Power BI implementation cost?

There is no single price tag for Power BI. This article breaks down what drives the cost of an implementation, what typical project scopes look like, and how to keep the budget under control.

· 4 min read