>As for good vs. evil, I believe that most people are good. Therefore I think that most bots, being deployed by humans and not yet being intelligent in their own right, are either good or benign.

The problem with that logic:

>Of course, people with nefarious intentions could be deploying more bots than good or benign people.

Is precisely that.

There can be one bad person for every thousand good people, but one person could automate countless "evil" bots. Yes people could deploy good or benign chat bots, but if someone wanted to troll or spread misinformation they would just deploy an army of chat bots across a wide scope of the social media.

Anyway, I'm not defining good or evil here, just going along with those words to keep it simple. Evil in this situation can refer to any form of deception from advertising to hate speech. If the bar for evil is simply not disclosing that it's a chat bot, I think that brings money and political gain into the mix which closing the gap of good vs bad people.