If you just say, "here is what llm said" if that turns out to be nonsense you can say something like, "I was just passing along the llm response, not my own opinion"
But if you take the llm response and present it as your own, at least there is slightly more ownership over the opinion.
This is kind of splitting hairs but hopefully it makes people actually read the response themselves before posting it.
"People are responsible for the comments that they post no matter how they wrote them. If you use tools (AI or otherwise) to help you make a comment, that responsibility does not go away"
People will still do it, but now they're doing it intentionally in a context where they know it's against the guidelines, which is a whole different situation. Staying up late to argue the point (and thus add noise) is obviously not going to work.
I'd prefer the guideline to allow machine translation, though, even when done with a chatbot. If you are using a chatbot intentionally with the purpose of translating your thoughts, that's a very different comment than spewing out the output from a prompt about the topic. There's some gray area where they fuzz together, but in my experience they're still very different. (Even though the translated ones set off all the alarm bells in terms of style, formatting, and phrasing.)