Update to Rule 13 – AI-Generated Writing > Likes and Comments
date
newest »
newest »
"Moderators may use AI-detection tools" AI- detection tools are AI.
Bill wrote: "...AI- detection tools are AI."Yes they are. And are known for errors. They're only as "good" as their LLM - and there's no telling what they were trained on. (They too often don't say.)
Yeah, respectfully, I am not a big fan of using this or worrying about it. I don't use AI for my writing, but I can see where this would be just another barrier and headache for the group and the moderators. Just my humble 2 cents. Thanks for what you guys do, though. I appreciate you.
For what its worth, I approve of this measure. It's one of the reasons I lean into humor in my stories. I figure it will be one of the last things AI does well.
Bill wrote: ""Moderators may use AI-detection tools" AI- detection tools are AI."We are well aware of that.
There are appropriate uses for AI, and inappropriate uses for AI. Police use facial recognition to search through thousands of faces when human eyes would take much longer. Doctors use AI to help them find possible cancer. Those are both appropriate uses for AI. Dishonest students use AI to generate their essays, then turn them in expecting a grade for work they did not do. This is an inappropriate use. Inappropriate uses utilize deception. Appropriate uses do not.
Please note this paragraph as well: Gene wrote: "We're aware of the limitations of AI detection techniques, and recognize there's at least a small possibility legitimate work might be flagged. If a submission raises concerns--whether through AI detection, well-known AI writing patterns, or other factors--we will review it carefully and fairly before making a decision. Our goal is simply to uphold the standards of the group while treating every author with respect and giving the benefit of reasonable doubt whenever appropriate."
I think the best way to prove you haven't used generative AI in your writing is to be able to show your writing process. If anyone wishes to question my writing, I can provide over 330 Word docs showing how I wrote HEAD and over 200 so far for my next novel!Any takers? Didn't think so! Hence the use of AI detectors, which I'm sure will become super accurate.
Gene wrote: "...For books first published in 2023 or later, authors will be asked to provide a PDF or Word copy of their book when submitting it to the group. This will be via the online sign-up form...."Questions...
If it's the same book that we've entered into a previous round, and it's been "cleared" for that previous round, do we then still have to provide a copy for new rounds?
And, if we're providing the mod upfront with a PDF copy of the book, will that eliminate the need to share our email addresses with everyone in the round, i.e. the mod can just send the copy to any of our reviewers who need a review copy?
Since there have been recent accusations of AI books in the rounds, books from 2023 and later will have to be cleared when signing up, even if they've appeared in previous rounds. Once your book has been cleared, you could state that in the sign-up form (providing a link to the previous round it was in would be appreciated).Sending out the review copies to the other reviewers will remain the same. I imagine most of the mods will be deleting the copies after we check them for AI.
Doesn't Amazon allow AI books to be published on their site anyway? Lol. I thought only AI reviews were in violation of Amazon's terms. I'm just surprised anyone is super worried about this from our end. It seems random and like another hurdle for moderators and everyone else to have to jump through, for no real value or return. I'm not trying to be rude. I just find it all kind of strange. Like, what is the point of worrying about all this, when Amazon itself doesn't even care about AI usage? Am I missing something? Or is this just a moral stand about the issue in general? I guess everyone has made up their minds, so we can see how it goes, but I don't even understand the point of wasting time on this, when Amazon doesn't even care. You can still leave a review about the quality of a book's writing, putting whatever you think about it, whether it seems like AI or human written.
The problem is people passing off the work of AI as human created. Even Amazon requires disclosure of AI written content.
Bill, your comment was removed from this thread, due to lack of civility. Please refrain from personal attacks or hostility.While we welcome comments, they need to remain civil.
For the vast majority of people participating, the only change will be adding a link to the sign-up form. If you're in good standing now, I can't imagine you won't stay that way. There also isn't much preventing someone from spamming the board with AI content now, most likely someone registering a new account, so we'll have to navigate a way forward. This isn't going to be a witch hunt. We're not looking to ensnare an innocent em-dash enthusiast. Moderators are following the same rules as anyone else.
The rules have always allowed AI book covers and interior art. That hasn't changed. Rule 13 has always specified AI text only, and enough AI patterns exist now to do some filtering before a round begins. That's all this is.
Gene is correct. Moderators are held to the same standards as others, and the text of books we have written, published 2023 or later, will be held to the same accountability.
Gene wrote: "...If you've listed a book in the One for One Review List thread that features AI text, you'll also need to have it removed, or swapped out with a book that doesn't feature AI writing..."Another couple of questions...
If I want to swap my book listed in the One for One Review List, will I need to provide the Mods with a PDF copy of my (new) book?
And if I don't yet have a book on the list and read and review a book there, will I need to provide the Mods with a PDF copy of the book I want to add to the list in order for it to be added?
Lynn wrote: "Gene wrote: "...If you've listed a book in the One for One Review List thread that features AI text, you'll also need to have it removed, or swapped out with a book that doesn't feature AI writing...."Hi Lynn,
Thanks for the questions.
If you're swapping your current book for a new book, the new book would need to be submitted to the moderators for screening before it's added to the list.
Likewise, if you don't currently have a book on the list but read and review one from the list and then want to add your own book, that book would also need to go through the same screening process.
Books already on the One to One List don't need to be resubmitted at this stage. However, if a book is subsequently found to contain AI-generated writing, including through screening in a Reading Round, it may be removed from the list in accordance with Rule 13. The author would then be invited to submit an alternative book that doesn't contain AI-generated writing.
I've just updated the One to One List to reflect the above.
Seth wrote: "Doesn't Amazon allow AI books to be published on their site anyway? Lol. I thought only AI reviews were in violation of Amazon's terms. I'm just surprised anyone is super worried about this from ou..."I think I am the newest mod, and I use AI to generate my covers--but then I put about 20-40 hours into editing them, and I am transparent about that on my copyright page as well as with Amazon when I am setting up my books. I have tried to stay out of this debate both behind the scenes and here because of that (even though AI-generated covers are not prohibited in this group) and because of my personal feelings about AI and data centers, in general.
That said, if you think that AI-generated work is harmless, I want to disclose that I have a lot of personal beef with AI. I was a freelance writer, and I was making a decent income until AI stole my job--I can't compete with free editing, translating, or ghostwriting. I can't even find jobs anymore fixing poorly written AI-generated books because people have stopped caring about quality.
Now, I get to write my own books, which doesn't really pay at this point, but I enjoy it more. Still, AI-generated books that were written with little human input are already taking buyers away from human-authored books and making it even more difficult for indie authors to succeed.
There are also several inherent problems with AI-generated books. For example, AI-generated work cannot be copyrighted as it stands right now. Arguably, if you change it 70%, you might have a claim, but that has not been established and would require long, expensive court cases. People still claim they own the "copyright" of these works when they do not. AI-generated work also tricks people. I have no doubt some of the people who use it think they are changing it and "making it theirs" more than what they actually are. Not only have I seen this firsthand, but I also understand and have seen what happens to beginning non-fiction writers who think they are paraphrasing something when they are actually plagiarizing it. Probably most importantly for this group is that AI-generated books are difficult to read because they frequently don't make sense. A bad writer can write as poorly as AI, but there are specific, especially irritating nuances that AI tends to prefer.
**All of the moderators have been extensively studying these AI nuances and debating them together for weeks.**
We have already had our suspicions about some books, but we did not jump to conclusions on this. We have just seen it building here. Some of the books you review in every review group may not be your cup of tea, but when you have to read one or two every round that don't make sense, have serious plotting issues, and are filled with many meaningless statements and literary devices, it becomes painful to finish them. And, yes, I can say for certain some reviews I have read recently are also AI-generated, too.
Initially, we as moderators decided not to police this, in part because of the amount of work involved, but most of you are extremely gracious, and no one has yet complained to me about having to read an AI-generated book or receiving an AI-generated review, so I have sat back and let it slide. Do you expect those people who are being dishonest about this and submitting AI work here to deal fairly with the review of your book? Keep in mind that Amazon is ramping up its policing because laws are changing to require disclosure: https://www.cnbc.com/2026/07/23/amazo...
It is my understanding that when you submit your post-2023 manuscript to us, we are going to read it first. We (humans) are going to make the initial decision as to whether or not we believe it is AI-generated. If we decide there is an issue with it, we will submit it to a machine to double-check our opinion. Based on recent studies, it is more likely that we (and the AI-checker) will label something that was written by AI as a human-written book than the reverse: https://www.forbes.com/sites/dereknew...
I, personally, have not reviewed any of the other moderators because two of them have reviewed me, and I don't think I have been in a group with E.K., yet. However, I have read some of their books. Even though I can't officially review them, I don't suspect any of them are using AI. However, I will still check their work if they join one of my groups and submit a book written post-2023. Obviously, if I am the moderator of a group, I won't suspect my own work since I know I didn't use AI to write it, but if anyone in my group told me s/he suspected my work, I would gladly submit it to an AI-checker and share a screenshot of the results with him/her. (I use MS Word, so I don't keep a record of all my revisions.)
In fact, if I were moderating a group prior to this and someone had questioned any book, I would have reviewed it and submitted it to an AI-checker. Then I would have shared the results with the person who questioned it (if the book were not AI-generated) or notified the other moderators and followed through with the consequences above (if it were).
AI-generated work is a real problem, but this decision is not one that the moderators entered flippantly.
If I may,Some authors using AI in the creative process of writing fiction seem incapable of understanding these issues:
'Some of the books you review...that don't make sense, have serious plotting issues, and are filled with many meaningless statements and literary devices...'
It might be helpful to provide short paragraphs with examples for each of these.
To be clear on my point of view, I've read quite a few books that read more as 'product' rather than creative content in the past and have occasionally awarded them 4*.
My argument is that they passed the creative threshold to a 4* level because they had been edited to include a super distinctive human voice and author identity, and were creative enough to get a 4* review from me.
I have never yet read a novel with generative AI content that I would award 5 stars. Is it, therefore, only at the 5* level of human creativity that AI cannot compete?
I'm not going to lean into the arguments about who thinks a book is what star level. I'll stick with my own opinion on that and add that I review creative works primarily on creative concept, ability to refer to the literary canon, and then style/method.
To help enforce Rule 13 and maintain the group's focus on human-authored books, we are introducing an additional screening step for new submissions.
For books first published in 2023 or later, authors will be asked to provide a PDF or Word copy of their book when submitting it to the group. This will be via the online sign-up form.
Moderators may use AI-detection tools alongside their own assessment when reviewing submissions. If a submission raises concerns, it will be discussed by the moderation team before any action is taken.
We're aware of the limitations of AI detection techniques, and recognize there's at least a small possibility legitimate work might be flagged. If a submission raises concerns--whether through AI detection, well-known AI writing patterns, or other factors--we will review it carefully and fairly before making a decision. Our goal is simply to uphold the standards of the group while treating every author with respect and giving the benefit of reasonable doubt whenever appropriate.
Where the moderation team concludes that a book contains AI-generated writing, books may be declined and blacklisted. Repeated submission of books containing AI-generated writing may result in removal from the group.
If you have submitted a book you know contains AI writing in the past, please do the right thing and refrain from entering that book in a future review round.
If you've listed a book in the One for One Review List thread that features AI text, you'll also need to have it removed, or swapped out with a book that doesn't feature AI writing.