Depending on the size of the data load, if a loader tool isn't available and you're adding significantly more values than already exist in the table (eg, if you truncate and replace when you do your load), the indexing may slow down the import.
If you have this sort of situation, it may be better to drop the indexes, do the data load, then put back in the indexes. In some cases, if you don't trust the data being loaded, you may need to retain your unique indexes to verify that you don't have duplicated records.