‘Obvious markers of AI’: doubts raised over winner of short story prize
Granta publisher says ‘perhaps we never will know’ true authorship of work that won Commonwealth prize
A few syntactical tics – and the verdict of an AI detection platform – have sparked a furore over the possibility that a short story given a prestigious literary award was written by AI.
The foundation that awarded the prize and Granta, the magazine that published the winning story, said they had considered the allegations but had not reached a conclusion as to whether they were true.
“It may be that the judges have now awarded a prize to an instance of AI plagiarism – we don’t yet know, and perhaps we never will know,” the publisher of Granta, Sigrid Rausing, said.
The Serpent in the Grove was named as the winning entry for the Commonwealth prize from the Caribbean on Saturday and published in Granta magazine.
In “a voice of restraint and quiet authority”, according to the judging committee, it narrates an intense episode in a troubled marriage, and is set in a farmhouse next to an enchanted grove.
Shortly after it was published, internet sleuths – and a few literary critics – seized upon the work and its author, Jamir Nazir, reportedly a 61-year-old from Trinidad and Tobago with few publications to his name.
Ethan Mollick, a professor at the University of Pennsylvania, wrote on Bluesky: “100% AI generated story just won the Commonwealth prize for the Caribbean region,” calling it “a Turing test of sorts”. As evidence, he cited Pangram, an AI detector, which said the work was AI-generated, but also said: “Come on, if you know you know.”
Another commentator, previously employed at Palantir, said there were “plenty of other obvious markers of AI writing” in the story, including a litany of “not x, but y” sentence structures, by now a familiar trope.
Other pundits dug into what appeared to be Nazir’s LinkedIn profile, where he discusses matters including the AI arms race and AI replacing jobs.
The accusations are another episode in an ongoing, frenetic conversation about whether artists and creators are passing off AI-generated work as their own – and whether publications will be able to reliably catch them doing it.
The New York Times cut ties with a freelance journalist in March after he admitted to having used artificial intelligence to author a book reviewthat appeared to echo elements of one published in the Guardian.
The publisher Hachette cancelled the release of a debut horror novel, Shy Girl, over concerns it was written at least partially with AI.
Episodes such as these have fuelled discourse around the telltale signs of AI writing – words such as “delve”, a profusion of em dashes, and “vague, soft intensifiers” such as “quietly powerful” and “deeply transformative”.
They have also generated energetic business for a new cottage industry of AI detectors such as Pangram, which purport to be able to separate machine prose from human efforts.
Pangram performs well in controlled tests, but research into the efficacy of AI detectors predicts there will be “a continuous technical arms race” between the detectors, AI models and writers adapting their usage of them.
The Commonwealth Foundation and Granta have said there is a limit to their ability to detect whether the allegations around Nazir’s possible use of AI are true.
The foundation said it did not use AI checkers in its judging process because supplying unpublished work to them “would raise significant concerns surrounding consent and artistic ownership”.
It said all entrants to the prize had avowed that their submissions were their own work and “personally stated that no AI was used”, something it confirmed with “further consultation”. It added that AI checkers were “not unfailing and infallible”.
The foundation’s director general, Razmi Farook, said: “Until a sufficient tool or process to reliably detect the use of AI emerges that can also grapple with the challenges pertaining to working with unpublished fiction, the foundation and the Commonwealth short story prize must operate on the principle of trust.”
Granta emphasised that it did not have control over the winning stories but merely published them as part of an agreement with the Commonwealth Foundation. It said it put the winning story into the AI tool Claude, which equivocated on the work’s provenance, saying it was probably not pure AI but probably not an entirely human creation either.
“There is, however, a certain irony in the fact that beyond human hunches, AI itself is the most efficient tool we have for revealing what is AI-generated,” Rausing said. “Until the Commonwealth Foundation comes to a definite conclusion, we will keep these stories on our website.”
The Guardian approached Nazir for comment.
Explore more on these topicsShareReuse this contentСхожі новини
QTCinderella lands CAA representation amid growing mainstream success
