I don’t think this is true, depending on the task they can be extremely hard to spot. You especially don’t want to accuse a student of cheating using AI without very concrete evidence.

New York City, Los Angeles, Chicago, San Francisco, San Diego, Seattle, Washington D.C., …

I believe the original reason was that there were people downvoting trans positive content, and this let people be transphobic anonymously. (Conversely, transphobic content can easily be removed by moderators.)

And let’s be honest, despite what reddit people might say, people don’t only downvote bad faith content, they use the downvote as a “disagree” button.