It's not at all trivial, and there's a surprising amount of academic research on whether it's even possible. Take the most extreme case: a bunch of new account

It's not at all trivial, and there's a surprising amount of academic research on whether it's even possible. Take the most extreme case: a bunch of new accounts with long-numbered names posting identical political messages. It turns out that many of these have (unique!) real people behind them. They're not bots, they're coordinated people who are part of a campaign that's either grassroots (sign up and make ourselves heard!) or AstroTurf (I've been paid by a PR company to...). Are these excluded as bots? After all, a lot of PR accounts are paid to post. Many other 'bot' accounts have scheduled high-volume posting, but are directly operated by a single human; they may be spammers, but they aren't anything other than who they claim to be. What about political clubs brigading? If I get my mates to set up alts and frequently shitpost about politicians we dislike, is that a bot army, or privacy-friendly political activism? In the end, you just have to take a view and decide what you're prepared to accept.