The Caine Prize for African Writing has announced that an internal investigation found no evidence that Zimbabwean author Simbarashe Kundizeza used generative artificial intelligence to write his winning short story. Consequently, the prize organization has officially confirmed him as the recipient of the £10,000 ($13,000) award for the 2026 cycle.
Ellah Wakatama, chair of the prize’s board of trustees, addressed the controversy in a video posted on X, stating that the review was triggered by an article in the US publication The Atlantic. The article, written by journalist Vauhini Vara, claimed that AI-detection software named Pangram identified Kundizeza’s story, The Blue Fighter, as entirely generated by a machine. The publication of these allegations sparked significant debate on social media, with some critics suggesting that such accusations disproportionately target non-white writers.
Wakatama explained that the board adopted a “human-centred approach” to the inquiry. The review process included interviews with the author, the 2026 judging panel, and representatives from The Shallow Tales Review, where the story was initially published. Additionally, an independent cyber security and digital forensic investigator conducted a metadata analysis of the drafts written by the author.
From this detailed investigation, the Caine Prize concluded that the evidence was clear: the story was written by a human. “We chose to take a human-centred approach to the investigation… we feel the evidence is clear: The Blue Fighter was human written,” Wakatama said.
Kundizeza, who lives in Harare, had previously denied the allegations. He told the BBC that his inspiration came from hearing a poem by South African Vangile Gantsho at an event in Ghana the previous year. The poem described her father’s obsession with building fish tanks, a theme Kundizeza wove into a fictionalised narrative about his own family’s struggles during the economic downturn in Zimbabwe around 2002.
This controversy follows a year of disputes over the use of AI in literary competitions. Earlier this year, the Commonwealth Short Story Prize was awarded to Trinidadian writer Jamir Nazir despite similar AI allegations, while Canadian-Haitian author Thélyson Orélien withdrew from France’s prestigious Goncourt Prize amid accusations that his debut novel was machine-generated, a claim he denies.
Classic drama. One wrong article creates a mess that takes months to clean up professionally.
Did anyone ask the judges how they felt after the controversy? Their integrity might be questioned too.
I read the story and felt very human. Sometimes AI detectors are just as bad as coin flips.
£10k is a nice boost, but clearing his name is probably what matters most for his career long-term.
It worries me that these accusations often target non-white authors. We need to address that bias in criticism.
Surprised The Atlantic didn’t issue a correction. Reputational damage to the author was significant and unnecessary.
The metadata analysis is key. Hope other publishers stop using these flawed detectors to cancel writers without proof.
Good to see a proper investigation instead of just trusting black-box AI detection tools. Accuracy matters here.