I'm facing the same issue. However there is a workaround posted here: https://forums.databricks.com/questions/18254/unable-to-read-file-using-pandas.html

Basically read the csv using spark and then convert to pandas