While you work in Power BI using a live connection, there is no way to control how the data model looks or its structure within the Power BI Desktop because the data model is managed from the server side, such as SQL Server Analysis Services or Power BI datasets. It means that there is no option for hiding columns directly within Power BI for effective ways to restrict access to columns:
Update the Data Model on the Server: The best way is to proceed with an update of the underlying data model on the server. An administrator model developer can hide the column in the data model by setting the visibility property to hidden in tools like SQL Server Management Studio (SSMS) or Power BI Service. Therefore, it will not be exposed to users connecting to the dataset.
Use Row-Level Security (RLS): Row-level security is most often assigned to restrict access to rows, but it can also be nontraditional about denying access to a particular column. For example, the calculated measure could be written to return blank or null values for restricted users so that they can view the sensitive data depending on the roles assigned.
Build a New Dataset: New datasets should be built with only the required fields in case a live connection cannot be modified. Excluding the confidential columns from the very beginning prevents them from being exposed in the reporting layer.
Such modifications require working with the data model owner or administrator since they necessitate an update to the source data model or configuration on the server. These processes will ensure security and consistency for all users and reports at the column level.
