Yeah, I totally agree with that about getting a broken english comment than something that is generated by AI. What I have found is that AI seems to not stray too far from the standard, intro/main point/body/summarize main point format. Which makes it pretty obvious because most people don't naturally write like that unless it is an academic paper. Which can cause issues. For example, I had a parent approach me at our last school board meeting because one of her daughters college professors is refusing to grade her work because he thinks she used AI to complete it. She could just be a really good writer who follows that standard tried and true format and now "Turn it in" or whatever tool they are using is flagging it. I have a feeling if I wanted to, it likely wouldn't be too hard to write something that fools software into thinking AI wrote it.