After weeks of silence, Thélyson Orélien and his publisher have finally responded in detail to the allegations of plagiarism and AI use. However, the response isn’t likely to end the controversy surrounding the author and his latest book.

Many will recall three weeks ago when his latest book, C’était Ça ou Mourir (It Was Either That or Die), was pulled from consideration for the 2026 Prix du Roman Fnac literary prize. The reason was a pseudonymous X (formerly Twitter) account claiming that the book was written or assisted by generative AI. For evidence, the account posted screenshots of a Pangram report showing that the book was 100 % AI.

That pseudonymous account was later revealed to be operated by Samuel Fitoussi, a columnist at the conservative publication Le Figaro. This raised questions about whether the allegations were politically motivated, something that Fitoussi denied.

Orélien staunchly denied using AI in the book, and his publishers stood behind him and the book. However, the scandal began to slowly expand into some of his earlier work, namely a short story he had published in 2012 and articles he had written for Radio Canada.

As the additional allegations of plagiarism came out, Orélien and his publishers fell silent. He quietly ended his book tour and the story largely died down.

However, over the past few days, both Orélien and his Canadian publisher, Éditions du Boréal, have released statements. Unfortunately, neither of them is likely to put an end to the controversy surrounding the author and his book.

Orélien submitted his response to the allegations in a letter to Reuters. In it, he continued to deny that he used AI in his latest book, but admitted to some of his past plagiarisms. On those he said, “unacceptable mistakes” and that he was “just over 20” at the time with a great deal to learn about writing.

He went on to say that he deeply regretted his mistakes and offered “my heartfelt apologies without reservation.”

At nearly the same time, Éditions du Boréal released their own statement standing behind Orélien and the book, however, acknowledging that they could not verify the novel before the manuscript was submitted on September 12, 2025. This is difficult because, according to Orélien, he began work on the novel in 2017, well before generative AI was widely available, and had completed his first draft in 2019.

Obviously, these admissions are more likely to fan the flames of the controversy rather than subdue them. Not only has it drawn more attention to this story, but the new information doesn’t do anything to exonerate Orélien. If anything, it may make it work against him.

Previously, many, including myself, felt that Orélien had been the victim of a rush to judgment. He was pulled from the long list for the Prix du Roman Fnac prize on the basis of a pseudonymous post showing an AI report of the book. Though AI detection is improving and Pangram is one of the better detectors on the market, it is still far from definitive evidence by itself.

Now, however, it may be the only evidence, and that is a serious issue.

Right now, Orélien is in a very difficult position.

He claims to have started the novel in 2017 and finished a first draft of it in 2019. He submitted the manuscript to his publisher in September 2025. That’s an eight‑year gap between the start of the novel and its submission that, according to his publisher, they have no way to verify.

Given the timeline that’s involved, he doesn’t need to provide any significant evidence. He could show that the draft (or a significant portion of it) was uploaded to a cloud service, emailed to a friend or otherwise shared with a third party. All he has to do is prove that the document existed before November 2022 (when ChatGPT launched publicly), and that’s fairly definitive proof that he did not use AI to write the book.

According to the publisher, that proof doesn’t exist, or at least they were not able to find it.

This is where I would say we examine the credibility of the author. However, that has taken a serious hit as well, as he admitted to having committed plagiarism in the past. Yes, it was in his deep past, nearly 15 years ago, but he didn’t admit to it until after it became part of the public conversation.

As a result, there are possibly two scenarios here.

First, Orélien could be a serial plagiarist who started out copying other people’s work early in his career and moved on to using AI to write or significantly assist in writing his book.

Second, he is a writer who made some mistakes early in his career and learned from them. Now he faces unjust and politically motivated accusations of AI usage that he can’t completely refute due to how he operates.

On one hand, we can’t expect an author in 2017 to prepare for being accused of using a technology that didn’t exist then. On the other hand, it’s frustrating that there is no way to verify the book before September 2025.

The book is officially in limbo. There’s not enough evidence to say definitively how the book was written, and that could pose a very real problem for Orélien and his publishers.

For some, this might be easy to brush off. Does it matter if the book was written or assisted by generative AI if the book is good?

Obviously, many people will say that it matters a great deal. Many, if not most, readers value the human creativity that goes into a book and feel that an AI‑generated book is a fraud, regardless of perceived quality.

But what about those who don’t care? Why should the use or non‑use of AI matter to them?

There’s one simple reason: if the book was generated by AI, it doesn’t qualify for copyright protection, at least in the United States.

According to the United States Copyright Office, to qualify for copyright protection, a work must have a human author. We saw this with the monkey selfie case and the US Copyright Office has said this explicitly when it comes to AI-generated work.

If this is the case, then Orélien’s publishers will struggle to protect their book as, in many places, it is not protected by copyright and is functionally in the public domain. To be clear, I do not believe that there is enough evidence to draw that conclusion right now. I am not encouraging piracy of this, or any other, book. But, without a copyright, the publisher can’t stop others from downloading, selling or otherwise distributing the book.

For the publishers, that is what is at stake here. They either have the rights to a very lucrative book that is still in contention for various literary prizes, or they don’t own anything. It all hinges on a simple question that doesn’t have a clear answer: what was AI’s role in the creation of the book?

There’s simply no way to know definitively.

For me, this is a very frustrating case. Usually, after examining the evidence and hearing both sides of the story, it’s possible to draw some kind of conclusion. Even if there isn’t time for a full analysis, we can get to some kernel of truth.

But that’s not the case here.

A single AI detection tool is not enough by itself to prove that a work was generated by AI. However, neither Orélien nor his publisher have done themselves any favors. Despite claims of the book being started in 2017, there’s no evidence of the book before September 2025 and Orélien has admitted to past plagiarisms.

While I think it was noble of Orélien and his publisher to address these issues and admit publicly to the issues that exist. However, from a purely practical standpoint, I have to worry that it was unwise.

They not only re-raised the controversy surrounding the book, but the new information only does more to harm Orélien’s case. Though a lack of evidence is not evidence of guilt, it does nothing to clear his name or explain the history of the book.

In the end, not much changes. We still don’t know how the book was written. We don’t have solid evidence to point to either conclusion. The book, in my mind, is in limbo. That’s exactly where it was three weeks ago.