(Kay DiBianca is currently on hiatus)
Over at Authors Guild:
“The model clauses below cover important aspects of AI uses of author’s works: specifically, prohibiting AI use of an author’s work without the author’s consent; licensing specific AI uses as subsidiary rights with fair compensation; protecting audiobook and translation rights from AI uses without the author’s approval; and governing the author’s permissible use of AI in submitted manuscripts as well as the publisher’s use of AI in connection with the work. Authors and agents may request that publishers use any of these clauses, and publishers are free to adopt them.”
Re: Authors:
“Author shall disclose to Publisher if any AI-generated text is included in the submitted manuscript, and may not include more than [a de minimis/5%] AI-generated text.”
Re: Publishes:
The Authors Guild is concerned about reports that some publishing professionals are uploading manuscripts and authors’ personal information into public, consumer-facing AI systems for uses such as generating summaries, assessments, and marketing copy without permission from the authors or adequate guardrails to ensure that the manuscripts are not used by AI companies for training.
Uploading or inputting a copyrighted work or an author’s personal information into public, consumer-facing AI systems without permission may constitute a violation of the author’s copyright or right of privacy, and it puts the author’s intellectual property and personal information at risk. Editors, agents, and others in the industry who have access to authors’ works should not upload their manuscript to or otherwise prompt consumer-facing chatbots with any author’s works without first getting the author’s written permission. Further, where consumer-facing chatbots are used in workflows, publishers and other industry professionals should ensure that they opt out of having the work used for training. All of the common chatbots provide this option. Publishers should also take care that any internal AI systems are sandboxed models with guardrails to prevent the manuscripts or author information from being used as inputs for training.
Publisher shall not upload the Work or any of Author’s personal information to consumer-facing AI systems for purposes such as generating summaries, assessments, or marketing copy without written permission from the author or as otherwise agreed to hereunder; and when such permission is granted, it shall ensure that the manuscript is not used by third-party AI companies for training, such as by opting out of allowing training in user settings.
And:
To prevent injecting any AI-generated text into an author’s work, publishers should not use AI to substantively edit manuscripts, with the exception of basic spelling and grammar- checking applications.
Publisher agrees and warrants that it will not use AI to substantially edit a manuscript (excepting the use of basic spelling and grammar-checking applications).
I have a friend who paid thousands of dollars to have her work, edited by publisher who put it in an AI program. When she received it back, there were so many errors that she had to rework the manuscript to correct them.
This was wrong on so many accounts, but I didn’t know her when she signed the contract, or I would have told her she was being a scammed.
And AI put the comma in after work, not me, and I didn’t catch it until after it posted…
Thousands! Scam is right! They are everywhere now, preying on naive writers.
Jim, thanks for sharing AG’s comprehensive recommendations. Whether publishers will adapt their contracts to meet them remains to be seen. Individual authors who push back and/or require limitations don’t have much leverage but AG is large enough to make publishers at least pay attention.
Another reason I’m glad to indie-publish. I make my own rules and set my own standards. And AI is not welcome in my books.
Wow, Jim. I saw the email from AG, but since I’m off on (most) weekends, didn’t get the chance to read it yet. I didn’t realize there were so many different ways for AI to sneak into our books. It’s disheartening to watch.
Claude Sonnet and Kimi Moderato can now do a superb job as context and/or developmental editor. I write the words, Kimi critiques for plausibility, continuity, plot integrity, character consistency, and story appeal. I get a thoroughly professional assessment, including recommendations for improvement, plot holes, the plusses and minuses, in a chat-friendly manner. My historical thriller draft stands at 97,000+ words and Kimi has handled extensive rewrites, whole plot changes, and supporting character deletions/additions.
To me, this is the way to use AI, not to lay down story, but to take the story that comes out of me and help it become a polished work. All the fluster about AI-generated text does not fit my use case and I’m not worried about my use of Kimi in this manner any more than if I had a paid-for editor on the phone working over my insertions and repairs.
My 2025 experience with AI text generation was that the quality waffled all over the place. Some AI agents are getting superb at following a story outline, a plot module, and brainstorming plot ideas. Yes, they can write whole chapters, whole books, but there’s always something goes off the rails. And the effort to set up the “story bible” to guide4 it takes as much time and thought as writing it in the first place. So I don’t use it that way even though I could.
For $20 a month I get an editor working with me, with words I write, Ideas I create, and a savvy sense of how I go about pushing my characters through thick and thin. It’s a whole lot quicker and more personal than swapping manuscripts in the mail and waiting, waiting for a reply.
My $.02.
Glad to hear AG has laid out guidelines for publishers and authors on AI. Thanks for sharing these, Jim.
Thanks for sharing this, Jim. So important for writers to know.
Late again!
Well worth the read & to save it.
Thanks for putting this together, Jim…
🤓
I can see a divide developing between books written in the classic manner and books generated by AI systems. Maybe they should even be sold that way. Some readers won’t care—they just want entertainment. Others enjoy the entire creative process and value the human element behind it.
If we want people to make informed purchases, it should be clear whether the concept and the body of the work were created by a human or by an AI.
Some authors use Grammarly, others use ProWritingAid. No one seems bothered by mechanizing that part of the process.
AI gives me a 24/7 editor that tolerates dumb or repetitive questions, nudges me to re‑examine sloppy work, remembers names, places, and facts better than I ever could, and never complains if I take a break to stretch or grab a snack.
It has never sulked or refused to work with me just because I reject its suggestions, and the free tier is pretty affordable.