We currently export email data--including the Subject Line--from Workfront as a CSV file. This file is then opened and edited in Excel, before being sent to an internal application (CGen). This export to CSV is done in Fusion.
In order for Excel to correctly recognize the non-Latin characters and render them properly, the file needs to be defined as UTF-8 (Comma Delimited). We can force this to work by opening the CSV in Notepad, manually inserting a uFEFF (zero-width non-breaking space, or BOM) as the first character, and saving the file. This gives Excel the information it needs to recognize the contents and render the non-Latin characters correctly. But there is no way in Fusion to force the CSV file to be created in a way which Excel understands.
Russell Howe worked with us on this. There was a change that needed to be made to Workfront as well; that has been completed (Link). This Fusion issue is still open, however.
Two files are attached, note that the BOM character is invisible in Notepad--but if you try to open these files in Excel you'll see that the BOM version works correctly. If you would like to see the actual files we're sending or the Fusion scenario we're using to send them please reach out to me directly.
