Two essays, one week, the same question: can a machine write a book?
A literary prize named for Pascal and a novel whose publisher admitted using generative AI have reopened, in two separate op-eds, the debate over what counts as authorship.
Two pieces published this week in the French press return, from different angles, to the same unease: what does it mean when a text is produced, in whole or in part, by artificial intelligence.
In Le Monde, the writer Jean de Saint-Cheron takes aim at the Prix Pascal, a literary prize bearing the name of the seventeenth-century philosopher, for having honored AI-generated work. According to his op-ed, the prize mistakes calculation for the making of art: a machine, he argues, feels no need to write, and the "author" of an AI-written book has not made the effort of going to find the work where it actually resides — within himself.
The same question, with a different case attached, runs through a piece in Bibliobs (Le Nouvel Obs). It revisits the affair surrounding the writer Thélyson Orélien, whose Canadian publisher announced that generative AI had been used in the making of his novel "C'était ça ou mourir". For Bibliobs, the episode exposes a deeper shift in how fiction is both produced and read: learning, or suspecting, that a text came from an AI changes what a reader believes they have read, even when the words on the page have not changed at all.
Neither piece offers a resolution. Saint-Cheron's objection is essentially philosophical, aimed at what a prize is supposed to reward. The Bibliobs piece is more a question about perception: why the imagined origin of a sentence seems to matter as much to readers as the sentence itself. Together they suggest that the debate over AI and literary authorship, already live in English-language publishing, has fully arrived in the French literary press — not as a settled controversy but as an open argument about where, exactly, a book comes from.
Sources
This article was written from these reports.