My experience with propaganda bots is that the really nasty hate stuff will usually be posted by actual, real people (perhaps as a result of being prodded by bot-provided outrage), and bots will rather have all kinds of more subtle hinting and agenda-pushing - because bots are managed by semi-professionals who care about bots not being blocked and (often) don't really care about the agenda they're pushing, while there also is a substantial minority of semi-crazy people who just don't care and will escalate from zero to Hitler in a few minutes.