No, as I said courts have been ruling the opposite. The act of training an AI is fair use. There have been cases where other acts of copyright violation may have occurred before getting to that step (for example, the download of pirated ebooks by Meta has been alleged and is going to trial) but the training itself is not a copyright violation.
You can argue about ethics separately but if you're going to invoke copyright then that's a question of law, not ethics.
Does that matter? There have been several major court cases at this point that have established that training an AI is fair use.
That's not reflected in the music billboards or the traffic going to various AI providers.
Or, it's a way to generate songs about whatever topic or in whatever style that you personally are interested in. That's what I use it for.
I've lately found that Copilot is really good at banging out simple Tampermonkey scripts.
I, too, am someone who's looking forward to trying this out.
Philosophically, people can always come up with differences to fret about. Philosophers have argued for millennia about things that are impossible to ever detect empirically.
Practically, no.
AI tools can be trained and run locally by individuals, not just by corporations.
Depends how well you fight against the prospect.
Whether it's a "legitimate complaint" is the community-dependent thing, I guess. Check the rules of the communities they're posting in. The instance being used by the bot account may also have rules about how people can use accounts there, I'm not sure how to report an account to their home instance but that seems like the sort of thing that should exist.
Chatting with fictional characters via AI is nothing new.