It's not like they need first to detect those users are bots.
They could use “fake = 1” to report the number of fake users. Or “fake != 1” to report the number of real users.
I can’t imagine the gymnastics they would have to do to make a fake user count palatable in a road show so again, I think your interpretation is most likely. Then again, I’m about a billion dollars short of my opinion mattering…
It's more likely to be some kind of prettified name like "internal public accounts for assuring control of external applications," some sort of technical name that doesn't really make sense by itself but is complicated enough that it looks like someone knew what they're doing when they came up with it.
Management would flip if they stumbled on anything titled "fake." It's way easier to just hide red flags behind technical sounding terms than it is to try to explain why a "problem" isn't a problem and shifting the blame to someone you won't miss.
Sure, they could be dumb enough to do that, but no evidence of it.
I agree, but all content _is_ generated by users.