Being a writer in 2026 is weird. I mean, it might be equally valid to say that being in 2026 is weird, period. But today I wanna talk about weird writer sh*t, so … yeah.
I will not bore you with a litany of all the reasons why choosing to publish my first proper book in 2026 was an ill-timed decision. If I could jump into a time machine – if, you know, I had one parked outside, ready to go – I would choose literally any other time to launch myself into the indie publishing sphere. 2019 would have been a great year … but I digress. Publishing’s never been an easy industry. And then genAI came along and threw another wrench into its already complicated works. It’s a topic I’m loath to touch because the internet is foaming at the mouth about it and I value my peace. For me the issue is pretty black and white – I have not and will not use genAI for any creative endeavors whatsoever – but I see no benefit and have no interest either in justifying my own opinion or changing other people’s opinions. But there are ripple effects that are impacting all writers, regardless of their personal stance, and that’s what I want to talk about.
A lot of the concerns I’m about to address have been percolating for a while, but they were starkly illustrated in the discourse around the recent Shy Girl brouhaha. For those of you who don’t know what I’m talking about, let me quickly bring you up to speed.
Shy Girl was originally a self-published horror novel by indie author Mia Ballard, who had other self-pub titles under her belt. [Full disclosure: I have not read any of Ballard’s books and my knowledge of her is limited to the information that was circulated in the wake of the Shy Girl debacle.] The book originally came out in early 2025, and was successful enough to attract the attention of Hachette, one of the Big 5 (traditional) publishers, who acquired the rights to it and proceeded to re-issue the book under one of their imprints – first in the UK in Nov 2025, with a US release planned for April 2026.
From the start, Shy Girl was dodged by allegations of genAI use. To be clear, this occurred prior to the Hachette acquisition but seemed to pick up speed last fall, especially after a popular Booktuber made a detailed video about the book and its issues. [I watched it and found it informative, though I am not here to co-sign its conclusions since, as noted, I have not read the book myself.] Fast forward to now: the New York Times published an article documenting its investigation of the allegations against Shy Girl, stating that (based on analysis using several AI detection tools) it found “recurring patterns characteristic of AI generated text, like gaps in logic, excessive use of melodramatic adjectives and an overreliance on the rule of three.” The next day, Hachette announced that it was cancelling the US release of Shy Girl and pulling it from shelves in the UK, citing its commitment to “original creative expression and storytelling.”
This is all by way of context. I am not here to re-argue the case for or against Shy Girl. Not my circus, not my monkeys. But, as an (indie) author, I can tell you that this case is the opposite of a victory for authors – and for “original creative expression”.
First and foremost, for me, it illustrates that traditional publishers are emperors with no clothes. They might hold themselves out as the (gate)keepers of professional writing standards, but Shy Girl shows that they will happily forego any vetting or quality assurance in the pursuit of a quick buck. This case also undermines the notion that traditional publishers are reliable partners for authors. They were happy to exploit Ballard’s popularity only to subsequently hang her out to dry when it became clear that she was more of a public liability than an asset. Again, this is not a question of whether the allegations against the book are true or not. It’s about the fact that Hachette acquired this book in the first place, only to turn around later and decide that it did not live up to its commitment to “original creative expression and storytelling.” Wasn’t it their job to figure that out beforehand? And if they’re not doing their job, what’s the benefit for an author of signing with a trad publisher?
Let me put it this way: it seems pretty clear Hachette did zero or minimal editing on this book, or else presumably they would have noticed its ‘questionable’ patterns long before now. That doesn’t exonerate Ballard if, indeed, she used genAI. But editing is one of the fundamental jobs of a traditional publisher; it’s why authors pursue trad pub, so they don’t have to pay for their own professional editing out of pocket. Forget AI for a second. A publisher who doesn’t edit a book – for basic grammar and formatting, if nothing else – is doing a disservice to both the author and the readers. If they are not doing their job, why are they getting paid?
Second, this case reinforces the public perception of self-published works as “slop” and, more pertinently, as “probably AI slop”, which hurts indie authors in several ways.
One, it encourages AI witch hunts and, at the same time, reinforces the idea that genAI usage can be reliably detected. Never mind that it can’t be. Never mind false positives. At this point in time, accusations of AI usage are pretty much career-ending events, especially for indie authors. The public might be willing to accept the likelihood of false allegations as a “necessary evil”, but for an author like me, it’s a terrifying prospect. There is still a TON of debate over the accuracy of AI detection tools like the ones used by The New York Times, and there are plenty of people who don’t even bother with tools and make accusations on the basis of things like em dash usage and general “vibes”.
Two, it makes readers less inclined to give indie authors a chance. I’ve seen countless comments online (in discussions of this case) from people saying they won’t read books from new authors written after 2021 or 2022 because of the potential use of genAI. Basically, all indie authors are being tagged as automatically suspect, regardless of their stated stance. You can imagine what it feels like to read that, as someone who is publishing books in 2026 and who is staunchly on the anti side.
Three, it de-incentivizes the disclosure of actual genAI use. Again, I’ve seen countless comments from pro-AI writers to the effect that they will continue to hide/not disclose their usage so as to circumvent the anti-AI sentiment of the reading public. This will only serve to reinforce the negative feedback loop of the previous issue. Ironically, it will also undoubtedly result in many AI-generated books flying under the “radar” while other, non-AI generated books will be cancelled for suspected AI use. Fun, fun, fun.
Coming on top of everything else, it’s enough to make one want to quit writing altogether. Well, quit publishing, at any rate.
All that being said, I have a lot of sympathy for readers too – I am one myself, after all. It’s a sh*tshow out there and one thing is clear: the corporations will not be riding in to save us from the mess. They’ll just make the mess, messier.
