Continuing the discussion from Discourse Version 1.2:
So I have 2 users appearing in this new “Suspect Users” list, but both of them are legit. What should I do?
AI-generated summary
The “Suspect Users” list in Discourse admin, introduced in version 1.2, flags users with low read time but a completed profile, potentially indicating spam accounts. However, some users, including downey and elberet, have reported legitimate users appearing on the list. cpradio suggests that users with Trust Level (TL) 1 or 2 will eventually be removed from the list, but downey points out that one of the flagged users is a TL2 moderator.
Suggestions to improve the list include excluding TL2 users or higher, as they typically require reading topics, and excluding staff members. downey proposes adding potential actions for flagged users, such as suspending or ignoring suspicion. codinghorror explains that suspect users can be removed from the list by actually reading content on the site.
Nothing. They should eventually get removed from that list once they make TL 1 or TL 2 (I think)…
It is purely a list of people you should “look into”, it isn’t going to inhibit them in any way.
One of these is a TL2 & a moderator.
Well that’s awkward 
Maybe the list can be improved upon then in some way to exclude those who are TL 2 (and it should definitely exclude staff)?
or perhaps it could be changed to let admins/mods dismiss them from the list selectively.
To me, it’d be nice if this list contained some potential actions for those that appear there … suspend user, ignore suspicion, etc.
I’ve had the system user on that list for ever, even tho it’s active, approved, TL4, admin and mod… 
lol, that’s funny! Don’t trust system, he is one shaddy character!
I suspect that it’s because I’ve deleted a few of its posts, like the “edit the first post in this topic to customize your FAQ/Guidelines/whatever” spam…
Suspect users have zero (or very near zero) read time but a completed profile. We found on many instances this correlates very highly with spam accounts.
One way to get off the list is to actually read content on the site!
I suppose not including them if they are TL2 or higher is also a good idea…
Granted seems a bit redundant as TL 2 typically requires reading topics… So maybe just excluding staff? I’m not sure how you moderate without reading topics either though…
Because sadly you have to be a moderator account for your user-based API key to do certain things (like granting badges, for one). Granted, a proper multi-API key permission-based system that isn’t user-bound would resolve that edge case.