Granta Literary Prize Accidentally Published a Story Created by Artificial Intelligence

The British literary magazine Granta, known for its prestigious Commonwealth Short Story Prize, which has existed since 2012, found itself at the center of an unusual scandal: one of the stories published this year appears to have been written by artificial intelligence.

A Story with Signs of Neural Networks

Jamira Nazir's text "The Serpent in the Grove" appears suspicious even upon first reading. It contains errors typical of LLM generation: strange mixed metaphors, excessive use of anaphora (repetition of constructions at the beginning of sentences), and an unusual abundance of bulleted lists. If traditional authors write from their own experience and intuition, neural networks often fall into formal syntactic patterns that over time become recognizable as machine fingerprints. The line between a well-trained authorial style and AI-generated text is becoming increasingly blurred. But experienced editors and critics are beginning to notice exactly these repeating patterns — unusual word combinations, logical oddities that no person would allow, but which LLMs generate with absolute confidence.

What Should Alert an Editor

What to look for when checking suspicious text:

• Excessive use of anaphora and parallel constructions — sounds almost musical, but appears too frequently
• Mixed or internally contradictory metaphors — LLMs take images from different contexts and combine them mechanically
• Unnecessary enumerations and sets of synonyms — characteristic "triads" and long lists that slow down the narrative
• Strange word choices in context — a synonym is technically correct, but sounds awkward and unnatural
• Absence of genuine emotional depth — text can be grammatically flawless, but spiritually empty

Of course, one sign means nothing. But when they stack together — that's a red flag.

The Trap of Trust in Literature

Granta is not just a publication, it's a cultural institution with heritage and authority. Its prize is a recognized standard in the world of literature. If an AI-generated text passed through the editorial process, blind judging, and made it into publication, current verification systems are simply inadequate for the new reality. The Granta story is a signal about an emerging problem. Sooner or later, every major literary magazine will face a similar situation. An important question: will they have the tools, protocols, and courage to quickly recognize such texts?

What This Means

The literary world is entering a new era where simply trusting authorship at face value is no longer sufficient. Magazines and prizes need new verification standards — from technical detectors to updated editorial guidelines and possibly requirements for disclosing the use of AI tools. For readers, this means that certification of "human authorship" could gradually become as important as the publisher's name.