A leading publisher has cancelled the US publication of a horror novel after claims that generative AI was used in its writing.
In what “appears to be the first commercial novel from a major publishing house to be pulled over evidence of AI use”, Hachette has blocked the US publication of “Shy Girl” and its UK edition has been discontinued, said The New York Times.
The “stunning fact” that the book got this far shows how AI is “seeping into even traditionally published fiction” and “how unprepared many in the book world are” for the “dawn of an uncertain new era”.
Article continues belowThe Week
Escape your echo chamber. Get the facts behind the news, plus analysis from multiple perspectives.
Sign up for The Week's Free Newsletters
From our morning news briefing to a weekly Good News Newsletter, get the best of The Week delivered directly to your inbox.
From our morning news briefing to a weekly Good News Newsletter, get the best of The Week delivered directly to your inbox.
‘Gaps in logic’
“Shy Girl” was originally self-published in February 2025, before being published in the UK in November. It was all set for a US release until The New York Times published claims of AI use.
Max Spero, founder of AI detection programme Pangram, ran a test that suggested 78% of the text was AI generated. The paper’s own analysis using several detection tools found “recurring patterns characteristic of AI generated text, like gaps in logic, excessive use of melodramatic adjectives and an over-reliance on the rule of three”.
Author Mia Ballard denies that she used AI and insists that an editor was responsible for the passages under scrutiny. “My name is ruined for something I didn’t even personally do,” she told The New York Times, while Hachette said it “remains committed to protecting original creative expression and storytelling”.
‘The plagiarism machine’
Everyone in publishing “knew a scandal like this would hit sooner or later” and “every editor I know has been crossing their fingers” that it wouldn’t be them, said author Lincoln Michel on his Counter Craft Substack. “More than a few” published books have been “partially or entirely written” by AI, but this fact has been “disclosed” and they used the technology in “thoughtful, artistic ways”.
The “layers of vetting and editing” used by traditional publishers are supposed to guarantee “a certain level of quality control” and “trust”, so they “may need to be a lot more careful now”. The episode may also make life harder for “emerging authors” because the “gatekeepers” of the industry will “have no choice but to figure out a way to drastically filter the flood” of AI, which might mean “leaning even more on connections” with established writers.
This “will not be the last time we see crap like this happen”, said Kayleigh Donaldson on political blog Pajiba. “More and more ‘authors’ will be exposed as users of the plagiarism machine”, but once a “big name writer” admits it there will be “no pushback” because they “make too much money”. Instead, there will be “smarmy think-pieces claiming that people are just jealous of AI and actually it’s sooo much better at writing than you are”.