Could a controversial award-winning short story signal a new era of AI 'literary slop'?
One of the winners of this year's prestigious Commonwealth Short Story Prize has been accused of using artificial intelligence to write his short story "Serpent in the Grove". Does the controversy signal a new era of AI "literary slop" – or merely a crisis in reading?
Issued on: 26/05/2026 - 11:53Modified: 26/05/2026 - 11:57
6 min Reading time Share By: Diya GUPTA
The Commonwealth Foundation announced the winners of its prestigious Short Story Prize on May 13. Five winning stories – one each from Africa, Asia, the USA and Canada, and the Caribbean and Pacific regions – were selected before the announcement of an overall winner. Aside from a small cash award, each writer's story is published on the website of famed London-based literary magazine Granta.
Granta has a long and storied history of publishing the early works of authors who eventually make their way into the literary canon. Sylvia Plath, EM Forster and Chimamanda Ngozi Adichie figure among a long list of acclaimed writers who were “launched” by the magazine. To most writers, being published in Granta and joining those distinguished ranks is a career-defining moment.
Granta is not itself involved in the selection process for the Commonwealth Short Story Prize, but lays the path for the winning writers to find their audience. The winner from the Caribbean region – a Trinidadian man Jamir Nazir – was one of the five selected from 7,806 entries this year.
Nazir’s story "Serpent in the Grove" may be among the most-discussed stories in Granta’s history – for all the wrong reasons.
Just days after it was released, readers noticed something odd about Nazir's writing. Postcolonial literature – from Derek Walcott to Jamaica Kincaid and Binyavanga Wainaina – is known for language play, but Nazir wrote lines that at best were a vague (“Her hair is midnight rain; her laugh is bright as zinc”, “Wilfred’s rum-shop leaned into the road like a rotten tooth”) and at worst, completely incomprehensible (“The girl smiled like sunrise over a sink”).
Accept Manage my choicesOne of your browser extensions seems to be blocking the video player from loading. To watch this content, you may need to disable it on this site.
Try again
Sharma Taylor, the judge for the Caribbean region, said the story was selected for Nazir’s “sublime” language – “…precise yet richly evocative – conjuring vivid, lush imagery with remarkable economy”.
But Granta’s readers disagreed. One researcher on X wrote, “Well, this is a first: a ChatGPT-generated story won a prestigious literary prize”, adding that the sentence construction and language were obvious markers of AI writing”. One reader said “the Granta AI thing reads like a literal parody of MFA lit” as another lamented that the “Commonwealth Prize has lost its credibility.”
Pangram, a company offering artificial intelligence detection tools, ultimately ran "Serpent in the Grove" through its systems, along with all the other winners. According to their results, 100 percent of the text was authored by AI. That wasn’t all: Pangram also said two other stories – Malta writer John Edward DeMicoli’s "The Bastion’s Shadow" and Indian writer Sharon Aruparayil’s "Mehendi Nights" were also deemed likely to have been written by AI.
Aruparayil denied the allegations, while Nazir and DeMicoli had not responded at the time of writing. It is worth noting that Pangram and other AI detection tools are not considered entirely accurate.
The Commonwealth Foundation’s statement said the judging process was “robust” and that the writers “personally stated that no AI was used”. Granta said they would take the allegations “seriously”.
Granta’s publisher Sigrid Rausing made the irony-laden decision to pass the story through Claude.ai, which concluded that "Serpent in the Grove" was "almost certainly not produced unaided by a human". (All generative AI – AI that creates new content based on patterns in data – requires human prompts.)
The waters have now been sullied. This literary scandal begs the questions: is AI changing writing, and how can we tell the difference between "human" and "human-ish"?
Is AI killing literature?
The Commonwealth Prize isn’t the first AI scandal in the literary world. Earlier this year, the Hachette Book Group withdrew horror novel "Shy Girl" after allegations that its author used artificial intelligence. Even Nobel Prize winner Olga Tokarczuk recently faced backlash after she admitted to using AI for research and drafting.
However, a handful of publications saw the AI tidal wave coming before it hit the shore. Neil Clarke, publisher and editor-in-chief of monthly science fiction and fantasy magazine Clarkesworld, cut off submissions after noticing a spike in AI-generated writing way back in 2023.
Clarke says he and his team now take a “hardline” approach to AI-generated writing. He laid out the process of selection and rejection in a blog post about changing the submission process, suggesting that the literary world isn’t quite dead (yet), it just needs to adapt.
Read morePrize-winning book 'Houris' brings prison term for French-Algerian author Kamel Daoud
“AI, for all its time-saving, has increased my workload by at least twenty-five percent,” said Clarke, who has a background in technology and computer science, as a large chunk of his time is spent sorting through "slop" submissions.
“From what I've seen, there are certain stylistic choices that AI seems to choose out of what it reads. The more you see these, the more you recognise that it's off.” Clarke has 17 years of experience publishing and reading all kinds of writing, and says that even if it’s difficult to know if a story is 100 percent AI, he says most “sticks out like a sore thumb”.
AI tools are not as sophisticated as their creators make them out to be. They’re prone to odd turns of phrase (as Nazir’s story exemplifies), hollow platitudes, and cheap imitations of human emotion – one example of which is OpenAI CEO Sam Altman’s "metafictional" literary short story about AI and grief which he said “got the vibe of metafiction so right”.
“These rollouts are not very good, they’re dishonest and misleading and full of mistakes. Notice that almost all the terms 'humanise' error – so when an LLM (large language models, an AI network trained on vast amounts of text) makes up a study, its creators call it a 'hallucination'. If they called it what it is, no one would wanna use it, because who wants to use a buggy piece of software?” says Clarke.
Clarke also uses AI detection tools – though he won’t say which, because then “the spammers would know how to get past it”.
“One of the things I keep pointing out to people is we've had spam filters for thirty years. They are invaluable, but they are not perfect – they have to be reviewed by humans,” says Clarke.
Accept Manage my choicesOne of your browser extensions seems to be blocking the video player from loading. To watch this content, you may need to disable it on this site.
Try again
But AI controversies in literature have done some irreversible damage, particularly when it comes to trust between writers, publishers and readers. “When you see so many stories coming that are generated artificially, it damages the way you look at things. If I read 10 AI stories in a row, I just feel disgusted by people – who does this? And then you approach a new writer with the same doubt and skepticism, and it’s not fair on them.”
Some writers have posited that the odd selection of this year’s Commonwealth winners may be less about a new war between people and machines – and more a simple crisis of reading.
Novelist Will Self, in an article on X he titled "The Novel Is Dead. This Time It's for Real", responded to the scandal with scathing criticism of Granta, which he said functioned “… less as a profitable enterprise than as a prestige object: a Potemkin village of literary seriousness”.
Self believes the crisis is "monumental": “The Granta affair is not simply an embarrassing editorial oversight. It is the first genuinely literary manifestation of a much larger civilisational fracture: institutions charged with preserving style can no longer reliably recognise it.”
As Booker Prize winner Marlon James wrote on social media: “Forget AI for a minute. A story won an International Competition with a line like this: 'The girl smiled like sunrise over a sink.'"
Advertising Keywords for this articleСхожі новини
Takaichi stresses ’great significance’ of retrial system reform
No Krishna Nagar in Lahore? Pak pauses heritage renaming plan after outrage by extremists
