Those sound like quite discretionary matters, including how the forum might treat the use of AI matter, since there are some members who have more egregious incidents than others, and they might even have other areas that they are breaking forum rules that justify more extreme sanctions against them.
With tags, if there were one or two incidents, but then the member might have had come clean or at least not had further incidents, then some members might choose to remove the tag, yet surely some members leave tags for extended periods of time too and might prefer to keep something like a neutral tag in order to serve as a reminder. Part of my own personal reason for not tagging members very frequently is that I feel that I would need to create another system for keeping track, which surely takes time to accomplish, but then the removal of tags might have some other reasoning besides what is written in the tag.. and surely we likely see that some members might have very shitty writing styles that don't seem to improve very much with the passage of time, and other members might improve in their substantive contributions to the forum and then justify the removal of the tag in the eyes of the person who put the tag in the first place..
I've sent quite a few neutral tags to people using AI in the forum. I wouldn't mind changing my initial trust rating on someone who was willing to come forward, apologize and promise to not use it again in the future. I'm all in for second chances, but you have to prove yourself that you're actually worth a shot. So far, no one has contacted me asking to change their rating, which is one of the main reasons I haven't changed it. I understand that a large percentage of the forum isn't native in English, including myself, but that doesn't excuse AI usage nor excessive usage of Grammarly/Quilbot, allowing the app to reconstruct your whole post itself.