Prize-Winning French Author Accused of Using AI; Furor Erupts
The questions the story raises are mostly not about the End of WritingIt was a perfect “happy ending” story for 2026. Quebec-based author, Thélyson Orélien, originally from Haiti, won the prestigious Prix du roman for a novel, C’était ça ou mourir (It was either that or die) about a migrant:
Fleeing Port-au-Prince with only a few poems as his possessions, Jonas Dorléon embarks on a visceral migration across the Americas to the United States. Between the harshness of the jungle and the humanity of his encounters, this novel recounts survival in a dazzlingly poetic language.
Orélien’s novel had already won the Méduse, the Première Plume and the Roman News prizes. It was then awarded the Fnac prize, sponsored by French booksellers.
That’s when the ax fell. Someone fed some electronic copy from the novel into Pangram, a reputable AI detector. At EuroNews, Nathan Joubioux and Alexander Kazakevich report,
The controversy was sparked by a message posted … on X on Monday 21 September by an anonymous account, “Balance ton Claude”. “According to Pangram, this novel was written almost entirely by an artificial intelligence,” it claims, saying it had tested several passages from the “manuscript” and almost always obtained “the same result: 100% AI”. (September 23, 2026)
As a result, It was either that or die (Grasset 2026) was dropped from consideration for the prestigious Goncourt Prize: “The Paris-based Academie Goncourt said it had taken note of the AI and plagiarism accusations, and had decided to remove his novel ‘to preserve the integrity of the Goncourt Prize and that of the Academie, whose central role is to promote and honour literature written by women and men.’” (Reuters)
The end of literature?
Some think it’s the end. British journalist and author Sean Thomas gloated at the Spectator that events have proven him right. People had angrily dismissed his 2023 claim that the End of Writing was upon us, thanks to AI.
Surveying the furore, he said, “Now, I’m not the sort of cheap, mean-minded person who says ‘I told you so,’ so I’ll just say ‘I told you so.’”
EuroNews quotes an AI enthusiast:
”I don’t understand what would be problematic about using AI to write a book in 2026. Not using it would be the stupid thing,” wrote … journalist Jérôme Colombain. “Ghostwriters have always existed. What matters, ultimately, is the author’s responsibility,” he added in the same post.
Colombain did not say why the ghostwriter, or for that matter, the AI, should not then get the Prize…
And then things got fuzzier
Orélien denied using AI:
Orélien spoke to France’s Liberation newspaper to deny the allegations.
“I’ve always loved creating and writing, I don’t see why I would use a machine,” the 38-year-old said, adding that he was “in the process of putting together a file” with his publishers to prove that he wrote the book himself.
“French-language author Orélien denies using AI to write award-winning book,” Radio France Internationale, September 24, 2026
He defended himself to the Canadian Broadcasting Corporation (CBC) as well: “‘I come from a Haitian and Caribbean literary tradition in which poetry, imagery, repetition, and rhythm carry a living voice,’ Orélien said in the statement written in French and released by Éditions du Boréal.”
But the picture soon grew complicated.
Reuters reports that he was accused of plagiarism on various occasions. RFI tells the same story. For example, from the Wall Street Journal, we learn, “An essay that won Orélien a prize from Radio Canada features large sections of text lifted verbatim from the preface of ‘Le Matin des magiciens,’ a 1960 book by Louis Pauwels and Jacques Bergier.”
The idea that the book might have been AI written had actually come up earlier. At WSJ, Anna Silman and Matthew Dalton note, “some readers online had criticized the book’s prose as overwrought, noting that parts sounded like AI, before the allegations erupted.”
Some analysts are alleging racism, claiming that black authors receive extra scrutiny. But then we must believe that Pangram is racist. Others think that Pangram misread Orélien’s Creole French. Or that AI detectors are generally unreliable, especially with second-language speakers or persons with special needs.
So now a big question looms
Why did this even have to happen?
Why had no one tried using an AI detector on Orélien’s—and everyone else’s—work as a condition of entry into the contest? Shouldn’t that be commonplace by now?
If a work tests positive, accusations of racism (or another prejudice) and of unreliable detectors will, of course, continue to be aired. But if AI composition is suspected at the entry point, there is time to defend the work, get independent evaluations, and/or learn something in the process. In any event, there is now an AI fiction writing contest, AI Pen, so flagged work can be directed there instead.
That way, AI writing won’t be a big uproar that casts a shadow over the entire current prize process or raises questions about the future of writing.
Yes, many questions remain. But some problems are comparatively easy to either solve or at least contain.
