Reported by Simon Daniel Yusuph l Journalist at Weng Global
Haitian-Canadian author Thélyson Orélien has denied using artificial intelligence to write his award-winning debut novel, C’était ça ou mourir, after an AI-detection tool reportedly classified portions of the book as entirely AI-generated.
The controversy emerged shortly after Orélien received the 2026 Prix du Roman Fnac, one of France’s major literary awards, for the novel. An anonymous account on X claimed that it had submitted passages of the book to Pangram, an AI-detection service, and received results indicating 100 per cent AI-generated content.
Orélien has rejected the accusation and said he has never used artificial intelligence to produce his writing. His publisher, Grasset, has also pushed back against the claims, while the controversy has raised broader questions about how reliably AI-detection systems can establish authorship.
Award-Winning Novel at Centre of AI Controversy
Orélien, a 38-year-old Quebec-based writer of Haitian origin, became one of the major literary discoveries of France’s 2026 publishing season with C’était ça ou mourir.
The novel tells the story of Jonas, a young Haitian history and geography teacher who is forced to leave Haiti because of violence and insecurity and embarks on a journey across the Americas.
The book was initially published in Quebec before its French-language European rights were acquired by Grasset. It was published in France on August 19, 2026.
The novel has attracted significant attention from readers and literary critics and has reportedly sold more than 35,000 copies. It has also appeared on early selections for several major French literary prizes, including the Goncourt, Goncourt des lycéens, Femina, Médicis and Renaudot.
The Prix du Roman Fnac was awarded to Orélien on September 21, adding another major distinction to the book’s rapidly growing profile.
How the AI Allegation Started
The controversy began on September 21 when an anonymous X account known as “Balance ton Claude” published claims that Orélien’s novel had been written largely with artificial intelligence.
The account said it had tested passages from the book using Pangram, a US-based AI-detection tool, and reported results showing 100 per cent AI-generated content in the material examined.
The allegation spread online at a particularly sensitive moment for Orélien because it came as his novel was receiving major literary recognition.
However, the existence of a high AI-detector score does not by itself establish who wrote a text or prove that generative AI was used during its creation.
That distinction has become central to the controversy.
Orélien Strongly Denies Using AI
Orélien has publicly rejected the allegation.
In comments reported by French media, the author said he had never used artificial intelligence to produce texts and questioned why he would need a machine to write when writing is central to his creative practice.
He also told the Montreal newspaper La Presse that the origins of the novel go back several years, with the work beginning in 2017 and an early version of the manuscript existing by 2019.
Orélien suggested that aspects of his writing style, including repetition and rhythm, could potentially explain why an automated system produced the results that triggered the controversy.
The author has also said that he cannot explain the detector’s findings.
His response means the central claim remains a disputed allegation rather than an established fact.
Publisher Rejects the Allegations
Orélien’s publisher, Grasset, has also defended the author.
The publishing house initially said it was not giving weight to what it described as rumours surrounding the novel. Other reports quoted the publisher as rejecting what it regarded as malicious claims circulating online.
No evidence publicly presented so far has independently established that Orélien used generative AI to write the novel.
The distinction is important because the accusation originated from an anonymous social-media account rather than from an official investigation, publishing record or documented admission by the author.
What Can an AI Detector Actually Prove?
The controversy has also drawn attention to the limitations of AI-detection technology.
AI detectors attempt to identify patterns associated with machine-generated writing. They do not observe the writing process itself and therefore cannot directly establish whether an author used an AI system.
Le Monde reported that it tested Pangram on more than half of Orélien’s novel and obtained an AI-generated score above 95 per cent, while the same tool classified the opening pages of a number of other novels as human-written.
At the same time, experts quoted in coverage of the controversy have cautioned against treating a 100 per cent detector score as proof that a text was written by an AI system.
A researcher interviewed by CelsaLab explained that a detector’s percentage should not be interpreted as a probability that artificial intelligence was used. According to the researcher, such tools identify linguistic patterns but cannot provide absolute proof of authorship.
That issue extends beyond Orélien’s case.
As generative AI becomes more common in education, publishing, journalism and creative industries, questions about how to distinguish human writing from machine-generated material are becoming increasingly important.
Why the Case Matters Beyond One Novel
The controversy arrives at a time when artificial intelligence is reshaping debates about authorship and originality.
Publishers, universities, news organisations and other institutions are increasingly developing policies around the use of generative AI. In creative writing, however, determining whether AI was involved can be considerably more difficult when there is no publicly documented record of the writing process.
The Orélien case illustrates the difficulty of relying solely on automated detection.
A detector may identify writing patterns that resemble machine-generated text, but such a result is different from documentary evidence showing that an author actually used an AI system.
That distinction is particularly significant when an accusation can affect an author’s reputation, literary awards, book sales and future publishing opportunities.
A Literary Career Under Scrutiny
Before the controversy, Orélien had emerged as one of the prominent new voices in French-language literature.
His background and the themes of C’était ça ou mourir have also attracted attention. The novel deals with migration, displacement, violence and the experiences of Haitians navigating difficult journeys.
The book’s success has placed Orélien in consideration for some of France’s most prominent literary prizes, making the AI allegations particularly consequential.
For the author, the controversy now sits alongside the recognition his debut novel has received from readers, booksellers and literary institutions.
The Prix du Roman Fnac award was based on a jury involving booksellers and readers, according to French media reports.
No Confirmed Evidence of AI Authorship
At present, the publicly available evidence does not establish as fact that C’était ça ou mourir was written by artificial intelligence.
There are competing claims.
An anonymous account has cited AI-detector results as evidence of suspected AI involvement. Some media outlets have independently tested portions of the novel and reported similarly high scores.
Orélien, however, denies using AI to produce the text, while his publisher has rejected the allegations.
Those positions cannot be treated as equivalent to definitive proof either way. The available evidence shows that a serious authorship controversy exists, but it does not independently establish that artificial intelligence wrote the novel.
The Wider Debate Over AI and Creative Work
The dispute also highlights a growing challenge for the publishing industry.
As AI-generated text becomes increasingly sophisticated, automated detection tools are likely to play a larger role in debates over authorship. But their results need to be interpreted carefully, particularly when they are used to make allegations against identifiable individuals.
For publishers and literary institutions, the issue raises questions about what evidence should be required before an author is accused of using AI and how such claims should be investigated.
For readers, it also demonstrates why an AI-detection score should not automatically be treated as proof of misconduct.
In Orélien’s case, the question remains unresolved publicly.
The author continues to deny the accusation, while the controversy surrounding his award-winning debut novel has opened a wider discussion about technology, authorship and the limits of automated detection.
Weng Global – stories beyond borders
Sources
Le Monde
Euronews
The Parisien
Livres Hebdo
RTL
CelsaLab
Journal de Québec
The Bookseller
