In the end, no one can prove if the accused user is using AI or not because these AI-detecting tools are not accurate, and not reliable at all like @KingsDen said as well. TBH, it will be hard to detect which of the posts here are being created by AI or not, and knowing that these AI detecting tools aren't reliable, it will take time before we know which posts are created by AI or not with high accuracy.
It's not hard to detect, there some telltale signs that someone is using AI for writing their posts, just look at the grammar, not a lot of people are using grammar checking apps to checks for grammatical errors in their post, although that's still not enough to determine if someone is using an AI. Another one that I can think of is trying to find if there's an emotion in the post, it's kind of weird but I think an AI generated post will sound more logical compared to someone who has written their own post, there's emotion in the latter because it's personal.
