Haitian-Canadian author denies AI claims after winning French literary prize
An award-winning debut novel by Haitian-Canadian author, Thélyson Orélien, has become the subject of an AI controversy, just weeks after earning one of France’s major literary prizes.
Orélien received the Prix du Roman Fnac for “C’était ça ou mourir,” translated as “It Was Either That or Die”,
The recognition has, however, been overshadowed by allegations that artificial intelligence was used to produce most of the novel.
The claims came from an X account, Balance ton Claude, which said it submitted different passages from the manuscript to the AI-detection platform Pangram.
“We tested several passages, from different bits of the manuscript, to come up with the identical result: 100 per cent AI”, the group said.
Orélien has rejected the allegation, insisting that the novel is his own work.
“I have always loved creation and literature. I can’t see why I would feel compelled to use a machine”, he told French newspaper Libération, according to the BBC.
He said he was working with his French and Canadian publishers to compile evidence that he wrote the book himself.
“My traditions are Haitian and Caribbean. The images, the repetitions and the rhythms are the living word, they are my own”, he said.
His French publisher, Grasset, also defended the author, calling the allegations “atrociously injurious” and describing the novel as original and powerful.
The controversy comes as the book continues to attract attention in France. It is also in contention for the Goncourt and Renaudot literary prizes.
The novel follows a teacher escaping gang violence in Haiti before making his way to Canada, where he rediscovers what the book describes as “the right to breathe without fear”.
Orélien was born in Gonaïves in 1988 and moved to Montreal in 2010, where he later worked as a journalist.
He said the first draft of the novel was completed in 2019, before generative AI writing tools became widely available.
The dispute has also reignited debate over how reliable AI-detection systems are in determining whether a piece of writing was generated by artificial intelligence.
