Image credit by ActualLitté. This file is licensed under the Creative Commons Attribution-Share Alike 2.0 Generic license.
The winner of the Prix Goncourt, one of the most prestigious and best known literary prizes in France, announced on November 4 that he used AI in his novel, L’Inconnue de quai de Javel (The Unknown Woman of the Quai de Javel). This comes after months of media attention directed towards another author, Thélyson Orélien, whose novel C’etait ça ou mourir (It was Either that or Die), about a teacher fleeing his home country, was removed from the prize’s shortlist, after accusations arose that he used AI to generate the text.
According to Le Monde, the prize-winner Phillipe Jaenada claimed that he used artificial intelligence to generate sentences. He admitted to using AI three times in his writing. Whilst the prize-winner has openly admitted to this, media outrage has largely been fielded towards Orélien.
The Canadian-Haitian author admitted to using AI in the research for his novel, something Jaenada also does, but strongly denies that any of his writing was machine-generated. Orélien’s novel was removed from the shortlist for the prize on September 25.
20minutes reports that from page 237 to 240 of Jaenada’s prize-winning novel, he describes an interaction with AI which he used to research the death of Louise Cansot, the real case which inspired his novel.
Jaenada transcribes the “absurd” answers of ChatGPT and Claude in the novel, commenting on them as part of the text. 20minutes also states that Jaenada mentioned his inclusion of an AI-generated sentence within the novel itself, as a challenge to the reader.
Despite this, in their press release L’Académie Goncourt argued that their decision to strike Orélien’s novel from the shortlist “affirms that [they] do not endorse or approve in any way the creation of texts generated with the help of AI.”
Moreover, after the claim about Orélien surfaced, the other 15 novels in contention were checked using the same detection software, Pangram. New Lines Magazine reports that according to the President of the Academy, the other novels passed the test, barring “the exception of a few ridiculous dregs.”
Just as concerning is the original source of the claim against Orélien: an initially anonymous X account under the alias Balance ton Claude (Call out your Claude). It later surfaced that one of the men behind the account was right-wing columnist and essayist Samuel Fitoussi, who initially denied his involvement.
Rather than taking the story to Le Figaro, one of France’s largest news outlets and a paper he works for as a columnist, Fitoussi’s collective chose to publish it anonymously on social media.
Literary critic Johan Faerber criticised Fitoussi’s journalistic method, as an anonymous X post bypasses conventions like the requirement of an editor, author impartiality and right of reply. The collective also didn’t check any of the other novels in contention for AI use.
Whilst AI-generated media poses a huge threat and a number of problems to the publishing industry, which should be taken seriously, the amount of media attention paid to Orélien’s case has been huge, regardless of whether or not he used AI to create his text.
Le Monde is one of the biggest news outlets in France. Searching the word “Orélien” on their online site yielded at least 22 published articles dealing with the topic in just over a month, as of November 7, half of which were within a period of 6 days (September 22 to 27).
By contrast two relevant articles published on the site appear after searching “Jaenada”, as of November 11: one from when he won the prize on November 3, and the other following two days later, reporting on his explicit statement to having used AI in his writing process the day after he won the prize. The latter article only appears on the English-language site.
Searching “Orélien” on franceinfo.fr as of November 7, another major news outlet in France, yielded around 55 articles, videos and audio files on the case published since September 21. Searching “Jaenada” produced at least one article on the topic of his “amusing” use of AI.
This comparison creates a number of questions. Where is the line for AI usage in the publishing industry? Can research be done using AI? Is it unacceptable to generate one or two sentences, or has a transgression only been made when a text is machine-generated in its entirety? More pressing: why is it okay for Jaenada to have openly used AI, when Orélien has been hounded by the media for accusations he continues to deny?
Orélien maintains that the first draft of his book was written in 2019, long before generative AI became publicly accessible to the degree it is today. He told Libération newspaper that his traditions “are Haitian and Caribbean” and some report that AI detection software is built around standardised language, rendering it potentially unable to analyse text using Caribbean dialect.
After cutting ties with him, Orélien’s publisher Boréal stated that they “unequivocally condemn the hateful messages and personal attacks of recent days, and [they] urge everyone to show restraint.” An X post from Balance ton Claude similarly points out an overreaction from the public sphere, calling for “a more measured approach to the attacks against Thélyson Orélien.”
In the original X post claiming to expose Orélien, the group Balance ton Claude claim that his work is up to 100% AI-generated. They say they felt morally obligated to expose Orélien, because “lying is morally reprehensible”.
Yet, they also implored writers to go ahead and use AI if they feel it adds value to their work, suggesting that it is acceptable as long as they mention artificial intelligence was used in the text’s creation. Jaenada claims that his use of AI is very minor and also intentional, something he did “to play with the reader”. Does this make it okay?
In a recent video posted by RTL France, Jaenada reveals the sentence he used AI to generate. He had it generate an idiom for him: “it doesn’t put bread on the table”, playing on its similarity to the French idiom “ca ne mange pas du pain” (it doesn’t cost a thing), literally “it doesn’t eat bread”.
The audience in the video responds to the reveal with applause and cheering. Yes, AI has no place in works of literature, but the difference in reaction to Orélien’s alleged use of AI and Jaenada’s confirmed use suggests a double standard, as to for whom it is acceptable to use AI in writing.
AI-generated work poses a threat to the publishing industry regardless of whether or not the author admits to its usage. If Orélien did write his book entirely using artificial intelligence, would it be acceptable as long as he had pre-warned his readers?
It is difficult not to compare the French media to the online response over the accusations of plagiarism against the late Professor Jason Arday this summer.
Professor Arday was the youngest black professor in the history of the University of Cambridge, and he was cleared of the accusations against him. The intense media scrutiny leveled against him at this time was often compared to a “lynching” after his tragic death in August.
249 articles were published surrounding the accusations against Professor Arday in a 22-day period. Professor Arday stated that he had “reached the limits of what any person should reasonably be expected to endure.”
Mental Health Europe reported that for years serious allegations, such as plagiarism, “against white academics were handled largely through institutional processes”, whilst Professor Arday’s case was made the focus of an unjust news investigation which probed into his personal life, and questioned why he was appointed to his position in the first place.
In the case of Orélien, the media has been accused online of replicating this racist double standard. Whilst article after article is produced on the accusations against Orélien, few major outlets seem to be interested in Jaenada’s use of generative AI, nor the implications thereof.
It is undeniable that the accusations against Orélien, true or false, have been weaponised to silence his voice. It is reported that he has since given up on his attempts to compile evidence proving his innocence.