Statement on Generative AI

The short of it is: fuck genAI.

The long of it is below, with thanks to Delilah Waan for her detailed example that she’s encouraged others to follow and thanks to Neil Clarke’s AI statement, which I agree with very much.

• generative AI tools should be based on datasets ethically obtained with appropriate compensation to and with the consent of rights holders
• datasets should be trained by workers compensated well, not exploited, and with sufficient support for their mental and emotional health
• governments should regulate usage to account for environmental damage and actively educate the population about the dangers to our climate (we did anti-smoking campaigns, we stopped using hairspray to heal the ozone layer; we can fucking do it here too)
• generative AI usage must be disclosed if a part of the creative or publishing process allowing reader to make informed decisions about their purchases, and generative AI should disqualify writers and artists from industry awards
• without the regulations accounting for environmental impact, protections for workers who train with data, and assurance of rights holders’ consent granted, there is no ethical way of using these tools

---

That’s my polite phrasing.

My personal opinion is that I enjoy writing, I enjoy revision, I enjoy creating photo manip covers and digital painting, I enjoy using my brain and building skills, and I believe in supporting human talent for anything I outsource. I agree with del Toro’s position and, while recognizing it’s being used not in the context of the current genAI discussions (therefore I cannot speak with authority to his position on it) I echo Miyazaki’s sentiment in that I find this slop an affront to life itself.

In terms of ethics, my opinion is that all of this based on stolen work (including mine), trained by exploited people in developing nations who have severe PTSD from sorting through graphic images and CSAM, and that is guzzling water and poisoning communities. The diversion of needed components for personal computers and medical equipment like RAM and hard drives being bought up for data centres and leaving the average person priced out of new hardware they often need for work is something everyone who has hyped this slop feel deeply ashamed of. There is no potential benefit to using it in the artistic process that can justify the enormous harm it causes. I am ethically vehemently against its use, whether by creatives, as a substitute search engine, or for the average person writing emails. I reject it on all fronts. I even block people permanently who visit my site from having made a ChatGPT query.

If my opinion is ever in doubt, remember that I have taken a huge blow to my career by choosing to no longer sell on Kindle due to their generative AI recaps (and now “ask the book” feature) and their unwillingness to add Terms & Conditions reflecting how they use our books with this technology.

---

Here is the breakdown of my publishing process, where generative AI could creep in, and how I ensure it doesn’t.

Brainstorming: Never. I take walks or take a shower or do dishes or talk to human beings. There is a fake name generator website I have used since well before the explosion of generative AI which I sometimes use for naming side characters (along with baby name websites and other non-AI sources).

Writing the book: I would rather give up writing. No book you buy from me or support on Patreon will ever involve generative AI tools.

Research: No. This is increasingly difficult in certain areas because of how many AI-generated articles and images there are out there. As much as possible, I’ve been relying on works written and published prior to 2022 when I can.

Editing and proofing: No—I work with real humans. I do run spellcheck/grammar check, I always have, but those are wrong more often than not now and I manually go through every suggestion to double check. I use Chicago Manual of Style and dictionaries. I am also way too cheap to pay for any other programs. I sometimes use text-to-speech as a final proofing step, having it read back to me to catch errors by the computer. TTS has existed long before current AI.

Generating jacket copy/summaries: Absolutely not.

Website: No. WordPress has been trying to shove it into everything but I refuse to use it (not on my main site, not on the Waverly Jones site).

Cover art: No. I do my own covers and this means I have to investigate every single image I choose. I’ve built up a huge stock library but I do still investigate photographers and their portfolios if I need something new. I have never generated an image using the AI built into Shutterstock or DepositPhotos for any of my covers (some previous tools, such as remove background, are now labeled under their AI umbrella, however; usually I don’t use that and extract models myself, but I have used it in the past). I use Photoshop but not the generative AI features.

Interior illustrations, maps, design elements: No, never. The odd time I hire an artist for something, I ensure they do not incorporate AI.

Marketing copy, website copy, social media advertising posts: Absolutely not.

Formatting: I manually do all my formatting in MS Word for print and produce epub files with Atlantis Word Processor. Any interior graphics were not generated by AI but use stock photos or handmade drawings.

Promotional graphics: Not knowingly. There are now a few I question from around 2021/2022 when unlabeled AI was being added to stock photo sites. Most of us had no idea at the time. I have never generated an AI image for promo, have always bought stock, and now I investigate anything new I buy.

Cross-promotion: I will not knowingly, willingly, cross-promote with any writer using generative AI in any part of their process. There may be certain circumstances–like if I’m in a Storybundle–where one of the other participating authors use generative AI. As I don’t curate the bundles, I can’t do anything about this.

Audiobooks: No fucking way. Either I find a way to afford real humans or I stop at four books available.

Translations: I would rather walk into the ocean than use an AI translation for a book.

How I Do Due Diligence

In terms of stock photos I use, I check upload date, check the photographer’s previous work, and look for any tells in the image.

There is zero chance I will use AI to write a book.

I sometimes use Canva and I’m careful about what graphics I choose; I do not use their generative AI capabilities.

Any artist I work with is one committed to not using AI.

Writers I cross-promote with are committed to not using AI.

Clients whose books I work on should not be using AI because I have made it clear to them I will not work with people who do.

Where Generative AI Could End Up Being Used

Although I make a reasonable effort, unscrupulous people have and will continue to upload genAI photos/images to stock sites and there remains a chance that I will accidentally use one.

There also remains a chance that, despite checking dates and confirming sources, at some point it could end up in my research material given the proliferation of usage in online articles.

Remedies for Inadvertent AI Usage

This happened, back in 2022, with background images used on one or two of my promos. I legit bought the image from a stock site and didn’t know it was AI until later. I stopped using the promos and deleted the images from my personal stock photo library once I knew.

If it’s brought to my attention that some element in my work used generative AI, I will investigate and replace the element by something human-made as quickly as possible.

If you believe generative AI was used in any part of my process, please feel free to contact me.

I will investigate anything brought to my attention.

---

Changelog: Initial statement December 27, 2025
Added emphasis and profanity to the opening February 17, 2026