In March 2025, Hachette Book Group pulled the horror novel Shy Girl from UK publication and cancelled its US release after a YouTube video and Reddit thread accused author Mia Ballard of using AI to generate the book. The incident is the second time in two years the publisher has faced AI-related allegations, exposing a problem for writers and publishers: the difficulty of distinguishing AI-generated text from human work, and the risk of being accused even when no AI was used.
Ballard denied the allegations, telling the New York Times she planned legal action against an editor she blamed for inserting AI content. Hachette had acquired the book after it was self-published with solid reviews. After the republication, scrutiny intensified. One Reddit user, a book editor, wrote: "If it isn't AI, she's a terrible writer. Her writing is truly indistinguishable from an LLM."
A second AI controversy for Hachette
In June 2023, freelance artist Robert Santora sued Hachette, alleging the publisher used AI to alter his book cover artwork for reprints of Sandra Brown novels. Hachette denied the claim, saying in-house designers created the covers. The case was dismissed because the works were not "substantially similar" to Santora's originals. Santora's complaint mentioned AI only as one possible method of creating derivative works, but media coverage focused heavily on the AI angle.
The lawsuit, though a routine copyright dispute, reinforced a pattern: even a publisher with a public anti-AI stance and multiple lawsuits against AI companies could not avoid accusations of AI use.
The difficulty of detecting AI writing
The cases reveal a dual challenge. First, publishers must screen submissions for undisclosed AI use, a task made harder by language models that change constantly. Second, writers who never use AI can still be accused based on superficial style markers. Wikipedia's 21,000-word guide on identifying AI writing lists em dashes, title case, and certain sentence structures as potential indicators. Yet many human writers have used those conventions for decades. As models change, the signals shift, making reliable detection almost impossible. Good human writing still stands out, but mediocre human prose and AI output often look alike.
Ballard's book had performed well during its self-published run. The accusations only surfaced after Hachette's involvement, suggesting that heightened visibility invites greater scrutiny. The publisher now faces the task of figuring out how AI-generated content, if present, made it through its editorial process.
Why this matters for writers
For writers, the Hachette incidents carry clear lessons. If your name is on the cover, you are accountable for every word inside, regardless of who contributed what. The line between human and AI writing is blurring, and even the appearance of using AI can damage a reputation or cancel a book deal. Keeping meticulous records of drafts and edits can help defend your process if challenged. While publishers work on screening tools, the burden of proving authenticity increasingly falls on the author. Writers concerned about these shifts can consult resources like AI for Writers Courses to better understand detection trends and protect their work.
Hachette must now determine how AI-generated content slipped through and build systems to prevent it, because similar incidents are inevitable. The publisher's anti-AI stance did not shield it from these controversies, and the same will hold true for writers who ignore the issue.
Your membership also unlocks:
