I've noticed there's a few different warnings. Like there's another one to do with sharing photo's of yourself with users you don't know.
That one always seems to be triggered automatically whenever someone sends me an opening message with certain keywords like 'image' 'picture' 'video' in it..
So I think there's a few scenarios. One is possibly the wording like if they'd used 'cash' 'credit' 'money' in their opening PM then it might be automated and the second is definitely when they're suspect.
I mean I've had new PMs before now which just said ''hello'' and that scam warning popped up on them. So it obviously can't always be written off as just automated keywords but like dirge said there's probably a few different metrics involved. 乂º◡º乂
