H.M. Wolfe's 'Daggermouth' Faces AI-Writing Accusations

A dystopian romance novel is testing what "AI slop" even means when the accused book has already cleared a seven-figure publishing deal and months on the USA Today bestseller list. The Atlantic reports that Daggermouth by indie author H. M. Wolfe, originally uploaded to Amazon as a Kindle ebook, tops a list compiled by academics of self-published Amazon books identified as containing "substantial AI text." None of the thousands of books they flagged has been more popular.

The specifics are quantifiable but tentative. Tuhin Chakrabarty, a computer science professor at Stony Brook University who researches AI writing, says the book has "too many telltale signs" of AI generation to buy Wolfe's account. The AI detector Pangram reportedly rated the manuscript around 60 percent AI-generated. Wolfe flatly denies it: "The suggestion that I used generative AI to write Daggermouth is wholly untrue," she said in a statement, adding that "accusations like these have real consequences for authors, and they should not be made on the basis of technology that is known to be unreliable." Simon & Schuster, which reportedly paid seven figures in February to secure the publishing rights to the book and its sequel, said through spokesperson Susannah Lawrence that Daggermouth went through "the same editorial and production process as our other published titles."

The awkward part for everyone involved is that detection tools are imprecise. Chakrabarty himself concedes that "if you significantly edit an A.I.-generated text, then it becomes human, and this is a harder problem in general." A book edited thoroughly with AI in the loop looks, to a detector, a lot like a book you wrote yourself. That means a major trade publisher is now defending a title against an accusation that, by the accuser's own admission, is hard to prove either way, while readers keep buying.

The honest caveat is that the reporting does not resolve whether Wolfe used AI, and no court, agency, or industry body has ruled on it. What the piece doesn't give you is a clear picture of what Simon & Schuster's editorial review actually consisted of, or what a genre publisher would even do differently if a manuscript came back with a high detector score.

The bigger shift is that if the market for commercial fiction is now pricing in books that may or may not have used AI at scale, and the tools meant to police that line cannot reliably tell, the argument stops being about "slop" and starts being about disclosure, contracts, and reader trust. That is a much messier fight, and it is where publishers, agents, and platforms are heading whether they want to be or not.