New Delhi: The Serpent in the Grove—a short story published in the UK-based Granta Magazine, and awarded the prestigious Commonwealth prize, has been accused of being AI-written.
This is the latest in the AI-writing allegations being made against popular works of literature. The Granta short story, written by Jamir Nazir—who is a 61-year-old man from Trinidad and Tobago—was flagged by literary critics, university professors, and X users because of the ‘obvious’ AI-writing syntax. The ChatGPT-style sentence structures, such as ‘not x, but y, ’ were found all over the story. New York Times Magazine, in its report, pointed out how the story is “crammed with metaphor and simile”. And also some odd descriptions: comparing a girl to “sunrise over a sink” and lines such as “She had the kind of walking that made benches become men.”
The story is about a troubled marriage, and according to the judging committee, it is written in “a voice of restraint and quiet authority”.
The problem with detecting AI writing
“There are other hallmarks of AI writing, like negative parallelisms and anaphora, or the repetition of words at the beginning of successive sentences or clauses,” the NYT magazine report said.
Most social media users accusing Nazir of using AI shared screenshots from Pangram, an AI-detection tool, as evidence. It claimed the text was 100 per cent AI-generated.
Nazir’s LinkedIn profile was dug up, where he had discussed issues such as AI replacing jobs and the AI arms race.
So far, Granta has acknowledged the allegations but has not concluded their validity, according to The Guardian.
“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, was quoted by The Guardian.
This case is part of an ongoing discourse over AI-writing and the anxieties about artists passing off AI-generated texts as their own work of art. The biggest issue, as flagged by readers and editors online, is that there is no sure-shot way to declare a piece of writing as AI-generated.
The Commonwealth Foundation didn’t use AI checkers before declaring the prize winner due to concerns about sharing unpublished work with generative AI. It “would raise significant concerns surrounding consent and artistic ownership”. Razmi Farook, the Commonwealth Foundation’s director general, told The Guardian that in the absence of a reliable tool to detect AI writing, the prize “must operate on the principle of trust.”
Recently, the New York Times fired a freelance journalist after he confessed to having used AI to write a book review. The article was found to be copied from a Guardian article. A debut horror novel, Shy Girl, was also cancelled by Hachette over concerns that it was partially written by AI, according to The Guardian.
Also read: One AI data centre generates more heat than 2 lakh households, US university study finds
The atmosphere of distrust around AI writing has spread across the world. In India last week, influencer and author Harnidh Kaur was accused on X of using LLMs to write her latest book, The Girls Are Not Fine: The Cost of Ambition, Careers and Becoming. Many called it “AI slop” and demanded that publisher Penguin Random House India recall all copies. Around the same time, IPS officer and columnist Pranav Jain faced similar allegations over his latest article in The Times of India on the “quiet grief of adult friendship”. Jain was called out on X even as readers widely shared lines from the column for how deeply they resonated. The piece reportedly got more than three million hits.
“A profoundly moving piece about friendships, that’s allegedly AI-generated, has managed to tug at readers’ heartstrings. Is AI getting better at verbalising raw human emotions or have human beings cracked prompting to get AI to emote this well?” a user wrote on X.