I'd venture to say Apple probably has a "less false positives" policy than the others. I can't say whether or not they do or do not scan, but if they review incidents with humans and not automated this could be why. They probably know flagging / disabling / reporting accounts incorrectly has a high cost on user satisfaction.