September 23, 2026, (Inside AI) — A literary scandal is testing the limits of AI detection tools and the trust underpinning the publishing industry. On Tuesday, the publishers of Haitian-Canadian author Thelyson Orelien's debut novel, C'etait Ca ou Mourir (It Was Either That or Die), publicly defended the writer against accusations that he used artificial intelligence to generate the prize-winning work.
The controversy erupted after an anonymous X account called Balance ton Claude claimed that an analysis by the AI detection tool Pangram indicated the novel was "written almost entirely by artificial intelligence." The accusation has since ignited a fierce debate in France over race, literary merit, and the reliability of the technology used to police originality.
Orelien, who was born in Haiti, has denied the claims. The novel, which tells the story of a Haitian migrant's journey north to Canada, has been longlisted for the prestigious Goncourt prize and has already won several other awards. The Académie Goncourt, which administers France's top literary honor, is now facing pressure to act before its next meeting on October 6, when it will establish a second shortlist.
"I did not expect attacks of this kind. I am not sad, but angry," Orelien told the French daily Libération. "I am in the process of putting together a dossier with my publishers in France and Quebec to prove that what I am saying is true."
Read: IIT Bombay Student Dies After ChatGPT Exam Cheating Incident, Protests Erupt
His publishers have echoed that anger and offered unequivocal support. Canadian publisher Éditions du Boréal stated, "We stand alongside Thelyson Orelien and thank the readers who continue to place their trust in him." His French publisher, Grasset, condemned what it called malicious claims, describing them as "deeply hurtful to a writer with an exceptional career." The publisher added, "The Grasset teams stand alongside this man ... they reaffirm their absolute confidence in him."
A Detection Tool Under Fire
The core of the dispute rests on the output of Pangram, a tool that claims 99.7% accuracy in identifying AI-generated text. However, the reliability of such software remains a subject of intense scrutiny. Independent experts have repeatedly warned that AI detectors can produce false positives, often flagging human writing that is formal, structured, or stylistically consistent.
This is not the first time an AI detection tool has been at the center of a high-profile dispute. In 2023, a similar controversy surrounded the use of AI detectors in academic settings, leading to wrongful accusations against students. The French literary establishment now finds itself grappling with the same technical uncertainty, but with far higher stakes.
The anonymous account behind the accusation has defended its actions on both technical and ideological grounds. In a chat on X, the account stated, "We don't see the connection between AI and skin color. Should we have stayed silent and allowed artificial intelligence to receive major literary prizes on the grounds that the official author is not white?" The account maintains that its anonymity is necessary to preserve its independence. Inside AI could not independently verify the identity or motivations of the individual or group operating the account.
The racial dimension of the attack has added a volatile layer to the debate. Orelien told Libération that he has faced racist attacks since arriving in France. The accusation of AI use, he suggests, is a new vector for an old prejudice. This claim resonates in a French publishing industry that has historically been criticized for its lack of diversity, particularly regarding authors from its former colonies.
The Goncourt's October Reckoning
Philippe Claudel, the president of the Académie Goncourt, has adopted a cautious stance. He confirmed that no decision has yet been made regarding the novel's place on the longlist. "If, by the time of our next meeting on October 6, when we establish the second shortlist, we have information that is indisputable, then of course we will have to consider the matter carefully," Claudel said.
Read: NYT Filings Reveal Microsoft, OpenAI Knew of 'Theft' Risks in AI Training
Claudel also acknowledged the broader implications of the case. He called it "almost inevitable" that questions over AI authorship are here to stay, and he advocated for a new framework. According to Claudel, a charter is needed that "makes it possible to distinguish what has been produced by AI from what has been produced by a human."
That call for a charter highlights a growing institutional anxiety. As generative AI models become more sophisticated, the line between human and machine authorship is blurring. Publishers and prize committees currently lack standardized protocols for verification. They are forced to rely on a patchwork of unreliable detection tools and the good faith of authors.
For Orelien, the immediate battle is personal and professional. He is assembling a dossier with his publishers to clear his name. The outcome will depend not only on the technical evidence but on the willingness of the literary establishment to trust an author over an algorithm. The decision by the Académie Goncourt on October 6 will be a critical test case for how the industry handles the next wave of AI-related controversies.
