If you’ve been keeping up with publishing news, you may have heard about Shy Girl by Mia Ballard, a novel that was recently canceled due to its purported use of generative AI.
While this post was prompted by Shy Girl, it is not another rant about generative AI (though I’m happy to write those as well!). Instead, I want to talk about the growing trend of traditional publishers acquiring self-published successes and rushing new editions to market. While this may seem like an obvious win for authors, publishers, and readers, what happened with Shy Girl exposes some of the serious underlying risks.
First: what happened with Shy Girl?
Shy Girl by Mia Ballard was originally self-published in February 2025 before being acquired by Hachette UK, followed by Hachette US. After the book was released by Hachette UK in November 2025 (and before a planned 2026 Hachette US release), accusations began to emerge online that the book had been written using generative AI.
As those in the book community know, these types of accusations are often ill-informed and can be notoriously hard to prove. Since generative AI models were trained directly on human-authored work (without compensating the authors), many of the purported ‘AI tells’ are legitimate writing techniques that humans have been using for decades, if not centuries (I personally will never surrender my beloved em dash!).
However, in the case of Shy Girl, the author’s response seems to confirm that generative AI was used—though by whom remains a point of contention. Ballard told The New York Times that “an acquaintance she hired to edit the self-published version of the novel had used A.I.” This statement itself prompts additional questions. To be clear: professional editors do not rewrite manuscripts or insert new sections of text, and they certainly don’t use generative AI to do so. Even the types of corrections made by copyeditors—such as adding a comma or capitalizing a name—are reviewed and approved by authors.
After Hachette canceled Shy Girl’s US publication and pulled it from shelves in the UK, the book community has been left wondering: how could changes of this magnitude be made to Shy Girl without the author’s involvement or approval? If the author had concerns about an acquaintance using generative AI, why didn’t she revert to a previous draft or share these concerns with her editor (especially given that some publishers have begun to contractually require authors to disclose any generative AI use)? Were there any signs of this generative AI use during the editorial or publication process?
Why is this an industry issue?
There are many reasons authors find traditional publishing deals appealing. Publishers have broad distribution channels; substantial budgets; large staffs of passionate people with expertise in editing, design, production, marketing, publicity, sales, subsidiary rights, and more; and many other resources that would be impossible for most individual authors to replicate. When these resources are thoughtfully deployed in support of a book, everyone connected to that book—from the author to the agent to the publisher to the reader—can benefit.
And when traditional publishers acquire self-published books like Shy Girl, their aim is to do exactly that: amplify the success and expand the readership that the author has already built. In other words, to pour gasoline on a fire originally kindled by the author. When this process works as intended, it can be great for all parties involved. There are numerous examples of books that were successfully self-published and that went on to even greater acclaim and success after being acquired by traditional publishers.
However, when a traditional publisher acquires a self-published book, they also create a ticking clock. Publishers naturally want their edition available to readers as soon as possible in order to avoid losing potential sales. Yet the typical acquisition-to-publication timeline takes over a year. If a successful self-published book was unavailable for that long, readers might move on. Momentum wanes, new books are published every day, and news cycles and attention spans are shorter than ever. Take too much time, and the original fire might go out.
The result can be immense pressure to compress the publishing process as much as possible. For example, a manuscript that might normally go through multiple rounds of in-depth developmental editing might receive a more cursory review; after all, why risk losing sales when readers have already responded to the original version? On the author’s side, they may be receiving detailed editorial notes and facing an aggressive delivery deadline for the first time. With unfamiliar pressures and a publishing deal on the line, it’s easy to see how generative AI’s promises of instantaneous results—or offers of help from scammers and charlatans—could become more tempting.
While I don’t know what happened during Shy Girl’s writing and revision process, it followed an increasingly-familiar compressed publication schedule: after being self-published in February 2025, the UK acquisition was announced in June 2025, just a handful of months before Hachette’s November 2025 UK publication date.
Who bears the brunt of the risk?
Though Hachette’s decisions about Shy Girl have certainly prompted questions, the cancellation of a single title is unlikely to have a significant impact on a multinational publishing conglomerate. The same cannot be said for Shy Girl’s author, who has faced immense personal criticism and whose publishing trajectory has been permanently impacted. Ultimately, it is the author’s name on the front of the book, and it is the author who suffers the most when things go awry in the publication process—whether due to their own actions or not.
I can’t help but wonder: what could have been done differently? If Shy Girl had benefited from a longer publishing timeline, would the use of generative AI have been caught before publication? Would the editorial team have been able to notice any inconsistencies across multiple drafts and conversations with the author? Would the author have been given the opportunity to address any concerns and revise the manuscript accordingly? In hindsight, all of these possibilities seem vastly preferable to Shy Girl’s high-profile cancellation—even if a lengthier publication process resulted in some lost sales.
Beyond the risk to any single book or individual author, the trend of traditional publishers acquiring successful self-published books poses risks for the industry as a whole. The more books that are rushed to market, the more opportunities there will be for similar situations to occur—with publishers, authors, and readers all paying the costs. And in an industry already fraught with inequity, traditional publishers could risk diminishing the very elements that should set them apart—finding and championing new, diverse voices; providing advances that allow authors to take the time to develop their craft; and ensuring thoughtful, incisive guidance at every stage of the publishing process.
From the outside, what happened with Shy Girl could be seen as a series of unfortunate mistakes; what happens next with the publishing industry will depend on whether it can learn from them.
Your Editor Friend,
Julie
P.S. If you enjoyed this post, please consider subscribing.