Ignoring RLS on certain visuals that are using the same dataset.

I am working with a dataset where Row-Level Security (RLS) is implemented, and we have multiple thin reports. We now have a requirement for one of these thin reports to display global numbers, effectively overriding the RLS implementation.

My current solution involves creating duplicate tables (both dimension and fact tables) specifically for this report to bypass RLS. However, this approach leads to having multiple fact tables with millions of rows, which significantly increases the dataset size and computational overhead.

Is there a more optimal way to achieve this without duplicating large fact tables? Any suggestions or best practices would be greatly appreciated.
My dax to filter the dataset looks like this.