In that case, the administration should take a position that would be clear, and if everyone (or the majority) shares your opinion, then it should not happen that some members, even after several dozen reported posts do not receive even a temporary ban due to the use of AI. If using AI inappropriately is the same as plagiarism, I guess the punishment should be the same.
I've seen quite a lot of tools wrongly detect a post as AI generated, when it actually isn't. I know you can say it has to pass through several tools before a conclusion can be reached, but if one tool is saying it isn't AI generated and another is saying it is, the accused user has an argument and banning them outrightly might not be possible. Using AI chatbot to post can only attract the same punishment as plagiarism, if it can be completely proven that the post is AI generated without any false positives.