Every user has their own view of what should or should not be abusive. HW's scope is clear and generally applied equally to each case. Of course, I take into account many different factors for blacklisting of downvoting, such as whether the user is new, intentionality, frequency of abuse, etc.

I do not blacklist new users. Neither do I downvote their new posts with abuse. New accounts that are downvoted are those that were evidently created as made-to-abuse.

If I see a new user who posted some AI spam or plagiarism, I would prefer to drop a general comment with guidelines, not specifying exactly what the abuse is. That's a different approach that I had in the past.