The Ghost in the Manuscript: The Growing Controversy Surrounding Thélyson Orélien

The literary world is currently gripped by a high-stakes investigation into the nature of authorship in the age of artificial intelligence. At the center of this storm is Haitian-Canadian author Thélyson Orélien and his latest novel, C’était ça ou Mourir (It Was Either That or Die). What began as a prestigious literary aspiration—a nomination for the 2026 Prix du Roman Fnac—has spiraled into a cautionary tale about digital forensics, professional credibility, and the existential crisis currently facing the publishing industry.

After weeks of strategic silence, both Orélien and his publisher, Éditions du Boréal, have issued formal responses to allegations that the novel was either entirely generated or heavily assisted by artificial intelligence. However, rather than extinguishing the fire, these statements have only served to expose deeper cracks in the narrative, leaving the book in a state of precarious limbo that threatens its commercial viability and its very status as a protected intellectual work.

A Chronology of the Controversy

The scandal erupted three weeks ago when the literary community was blindsided by the news that C’était ça ou Mourir had been abruptly pulled from consideration for the 2026 Prix du Roman Fnac. The catalyst for this disqualification was an anonymous post on X (formerly Twitter) that leveled a specific, damning accusation: the novel was the product of generative AI.

To support this claim, the accuser—later identified as Samuel Fitoussi, a columnist for the conservative publication Le Figaro—presented screenshots of a report from Pangram, a popular AI-detection tool. The report alleged that the text possessed a 100% AI-generated signature. While Fitoussi has denied that his investigation was politically motivated, the involvement of a prominent conservative journalist immediately infused the literary dispute with cultural tension.

As the scrutiny intensified, the scope of the investigation expanded. Critics began digging into Orélien’s back catalog, uncovering allegations of plagiarism in a short story published in 2012 and various articles he had authored for Radio Canada. Faced with a mounting crisis, Orélien abruptly canceled his book tour, and a tense silence settled over the author and his publisher. That silence was finally broken in mid-October, when the parties involved attempted to address the allegations, yet the resolution remained as elusive as ever.

Supporting Data and the Burden of Proof

At the heart of the dispute lies a frustrating evidentiary void. Orélien claims to have begun the novel in 2017, with a first draft completed by 2019. If true, this timeline would predate the public release of generative tools like ChatGPT by several years, effectively exonerating him of using modern AI to construct the narrative.

However, from a forensic standpoint, the claim remains unsubstantiated. When questioned, Éditions du Boréal admitted that they were unable to verify the existence of the manuscript prior to its submission in September 2025. This creates a "Schrödinger’s Novel" scenario: without a verifiable digital paper trail—such as cloud-saved timestamps, dated emails to editors, or metadata from early drafts—the author’s claims are effectively unverifiable.

For an author in the modern era, the absence of such evidence is a significant oversight, yet it is one that many writers are only just beginning to grapple with. In the legal and professional spheres, the burden of proof is shifting. While it may seem unfair to expect an author to have "proof-of-humanity" dating back to 2017, the lack of any traceable history leaves the public with little choice but to rely on the author’s word—a word that has been tarnished by his recent admission of past plagiarism.

Official Responses: Apologies and Deflections

In a letter to Reuters, Orélien broke his silence by addressing two distinct fronts: the AI allegations and his history of plagiarism. Regarding the latter, he offered a candid admission. He characterized his past instances of copying as "unacceptable mistakes" committed when he was "just over 20," emphasizing that he was a young, developing writer at the time. He offered his "heartfelt apologies without reservation" for these past transgressions.

However, when it came to the AI allegations, his stance remained firm and uncompromising. He vehemently denied using generative AI for C’était ça ou Mourir.

Simultaneously, Éditions du Boréal released a statement doubling down on their support for the author. While they stood by the quality and integrity of the book, their statement contained a troubling caveat: they conceded that they had no method to verify the provenance of the manuscript before it was submitted in September 2025. By admitting they could not verify the creation process, the publisher inadvertently highlighted the vulnerability of their own vetting process, signaling to the industry that the safeguards against AI-generated content remain largely performative.

The Copyright Crisis: Does the Book Exist?

Beyond the reputation of the author and the ethics of the publishing house, there is a profound legal question at stake: the copyrightability of the work itself.

In the United States, the Copyright Office has been consistent in its stance: copyright protection is reserved exclusively for works created by human beings. This precedent, established through various cases including the famous "monkey selfie" dispute, has been explicitly reinforced in the context of AI. If a book is determined to be the product of generative AI rather than human creative agency, it effectively loses its status as intellectual property.

This creates a nightmarish scenario for publishers. If a book is deemed to be AI-generated, it enters a gray zone where it may technically be considered part of the public domain. In such a case, the publisher would be unable to legally prevent unauthorized distribution, digital piracy, or commercial exploitation by third parties. For a house like Éditions du Boréal, this is not merely a reputation risk—it is a potential total loss of a financial asset. They are currently betting the farm on an author whose primary defense is an unverifiable timeline and whose credibility has been compromised by historical plagiarism.

Implications for the Future of Literature

The case of Thélyson Orélien serves as a bellwether for the future of the publishing industry. We are entering an era where the "humanity" of a text is no longer assumed; it must be proven.

For authors, this implies a need for a new standard of digital hygiene. Maintaining version histories, utilizing platforms that provide verifiable timestamps, and perhaps even engaging in "open" drafting processes may soon become the norm for those who wish to insulate themselves from the growing paranoia surrounding AI.

For the public, the case highlights the limitations of current detection tools. While software like Pangram is improving, it is not a judge, jury, or executioner. Relying solely on these tools to strip an author of their credentials can lead to unfair judgment, yet ignoring them in the face of suspicious output is equally dangerous.

Ultimately, the controversy surrounding C’était ça ou Mourir remains unresolved. We are left with two competing, yet equally unproven, narratives:

1. The Serial Offender: Orélien is a career plagiarist who has evolved from lifting text from human authors to utilizing the high-speed synthesis of AI.
2. The Unjustly Accused: A writer who made youthful mistakes is now being crucified by an algorithmic "trial by fire" that assumes guilt because the author cannot prove a negative.

As it stands, the book remains in a state of limbo. The debate has not moved forward, the truth remains obscured, and the literary world is left to wonder: if we cannot prove who wrote a book, does it still belong to the author, or has it become the first casualty of the automated age? For now, the only certainty is that the shadow of the machine will continue to loom over the written word, forcing readers, publishers, and writers alike to ask, "Is this human?" before they ever turn the first page.