French Literary Controversy Erupts Over Author’s Alleged AI Use

Summary

France’s literary world is debating whether award-winning author Thélyson Orélien used artificial intelligence to write his first novel, C’était ça ou mourir. The 38-year-old Canadian-Haitian writer recently won the Prix du Roman Fnac and is also a finalist for the Goncourt and Renaudot prizes. Balance ton Claude, a little-known group whose name refers to Anthropic’s Claude assistant, claimed that passages from different parts of the manuscript were “100 percent AI” according to the Pangram detection program. The group said Pangram’s error rate was almost negligible and later cited a rate of one incorrect result in 24,400 cases, but its membership and testing process are unclear. Orélien denied the allegation, saying the book’s Haitian and Caribbean imagery, repetitions and rhythms were his own, and said he was preparing evidence with his publisher. Grasset defended him and described the work as original and powerful. The novel follows a teacher escaping gang violence in Haiti before reaching Canada. The dispute echoes the case of Trinidadian writer Jamir Nazir, whose prize was withdrawn and later restored after organizers accepted that the writing was genuine. Critics and research cited in the report warn that AI detectors infer statistical patterns rather than identify authorship: they can falsely flag human writing, miss edited or paraphrased AI text, and struggle with dialect. Orélien also said the first draft was written in 2019, before text-generation tools became generally available. The controversy is therefore raising a broader question about what mixture of human writing, AI generation and AI-assisted editing should be acceptable in literature, without establishing whether Orélien used AI.