It always pays to try different AI detectors. Copyleaks is also my first choice, but try quillbot.

In this case, copyleaks says 39% of the post is AI (the first sentences were a dead giveaway). Quillbot identifies >60% as AI.

You also need to keep in mind that there is undetectable.ai, a service that claims to "humanize" the AI dreck that people can produce with ChatGPT & Co.

My way of dealing with them: If I detect AI in a comment to one of my posts, I check it, if it is AI, I also check out the commenter (I don't call them authors), and run some of their posts through the AI detector. In most of the cases, their posts are 100% AI. I then report several of their posts as AI to Medium, report their comment as AI to Medium, and block the "author".