@jesse_b The issue is when text is based on LLM output, and when it uses generative AI to just add words to make it seem authoritative. Using AI to fix grammar etc. is fine, in my opinion, and I would possibly also say that using it to check posts for other types of issues also is okay. It's not okay to let the AI be responsible for the content. You should be able to act on comments and to modify the text in response to issues.
Most users that we delete AI answers from don't even care and wouldn't be able to explain or fix their text even if they wanted to.