A PhD student at Southern Methodist University lost a publishing deal reportedly worth more than $2 million last month after allegations he used AI to write parts of the book.<\/p>\n
The author, Jerry Falade, denied the claims in a social media post.<\/p>\n
\u201cI\u2019ll talk when the time is right,\u201d Falade wrote. \u201cDefinitely innocent!!!\u201d<\/p>\n
The book\u2019s publishing rights were purchased by Minotaur Books in a competitive auction with 13 other publishers, according to The Bookseller<\/a>, a book publishing trade magazine.<\/p>\n Falade\u2019s book agent, Marc Gerald, told the magazine Falade gave conflicting answers about his use of AI after concerns were raised by another editor halfway through the book submission process.<\/p>\n SMU told KERA the book is unrelated to work done at the school, and declined to comment further. <\/p>\n KERA reached out to Falade, Gerald and Minotaur Books but did not hear back. This article will be updated with any response.<\/p>\n \u201cI\u2019m just a boy in my 20\u2019s,\u201d Falade wrote on social media. \u201cI don\u2019t understand why anyone is witch-hunting me.\u201d<\/p>\n The crime novel, \u201cCall Me, I\u2019ll Hide the Body,\u201d is the latest book backed by a large publisher to face accusations of AI use.<\/p>\n Hachette Book Group cancelled a planned horror novel, \u201cShy Girl,\u201d in March after the author was accused online<\/a> of using AI to write the book. \u201cDaggermouth,\u201d a popular dystopian romance novel, also faces AI accusations<\/a>, but its publisher has denied those claims.<\/p>\n Jane Friedman, who\u2019s worked in publishing for more than a decade and written a book on the industry, said publishers are responding to public backlash.<\/p>\n \u201cThere is this really strong moral hazard associated with using AI,\u201d Friedman said. \u201cYou have this piece of technology that is doing something that creators feel is sacred, which is creating expression, creating language, but there’s no human intent behind it.\u201d<\/p>\n There\u2019s also a legal concern.<\/p>\n The U.S. Copyright Office<\/a> released a report in 2025 outlining that AI-generated content can only be copyrighted when it has \u201csufficient expressive elements\u201d from a human. But that human element cannot be just the prompts written to the AI. <\/p>\n \u201cThere’s also a lot of gray area in here,” Friedman said. “We’re not going to know where the line is until there are court cases that examine it.” <\/p>\n The prevalence of AI, and the difficulty of accurately detecting it, has created paranoia between author, publisher and literary agent, she said. <\/p>\n \u201cWe\u2019re just operating in a world of both confusion and suspicion. No one really trusts anyone anymore,\u201d Friedman said. \u201cThere are some people who think, \u2018well, we can just work on the honor system.\u2019 I think the honor systems unfortunately are broken.\u201d<\/p>\n Falade told The Bookseller he was concerned his book was put through faulty AI-detection software. His agent denied that claim.<\/p>\n Friedman said there are no easy answers when it comes to detecting AI usage. <\/p>\n \u201cI have yet to see anyone come forward with a feasible process or solution that everyone is going to agree on,\u201d she said. \u201cI do think there’s probably going to be some sort of combination \u2014 and it’s going depend on the publisher \u2014 of using technical tools, human judgment and relationship building with the author.\u201d<\/p>\n Dylan Duke is KERA’s breaking news reporter. Got a tip? Email Dylan at\u00a0dduke@kera.org<\/a>.<\/p>\n