It seems that scammers who use AI to mimic actual human writers are getting smarter, after a British journalist found a memoir that bore a shocking resemblance to his own that he’d just published — except that this one was full of made-up stories and was published under a different name than his.
In an interview with The Guardian, former BBC tech reporter Rory Cellan-Jones said that when he went to Amazon to check his author bio, he found something he hadn’t written: a biography, penned by someone (or something) else, about his life.
“I thought: ‘This is strange — who’s writing a biography of me?'” Cellan-Jones told the paper. “It’s difficult enough for me to sell books about myself, [let alone] for other people to sell books about me.”
The book, published under the name “Steven Walryn,” was nominally a biography of the BBC alum, but turned out to be a “complete fantasy” that strongly appeared to be generated by AI. Its story diverged almost comically from the author’s actual book, “Ruskin Park: Sylvia, Me and the BBC,” detailing the sordid love affair between his single mother and his absentee father.
“There are passages about the Cellan-Joneses, an academic family sat around the table,” the true author said. “His father, a kindly academic; his mother, a teacher.”
“Just complete baloney,” Cellan-Jones quipped.
To make matters that much worse, the real Cellan-Jones told The Guardian that after he looked at the book online, he got an email from Amazon recommending the same phony biography about him instead of his own work.
“Amazon sent me an email saying: ‘You might like this,'” Cellan-Jones said. “Their algorithm had decided this was a bloody book I would want rather than recommending my book that I’ve slaved long and hard over… They’re effectively allowing book spam and recommending it to the very person who is most annoyed by it.”
Even more strangely, the profile for books purportedly written by “Steven Walryn” featured more than 30 titles that were a mix of repetitive, throwaway camera how-to guides and fantasy romances, with half of them having been published on a single day earlier this year. Amazon has since removed the titles, but as The Guardian‘s prior reporting on the issue shows, it’s easy for stuff to slip through the cracks until it’s brought to the e-commerce giant’s attention.
In August, the newspaper reported that Jane Friedman — whose work, ironically, focuses on how to publish books — discovered a bunch of phony titles under her own name that she suspects had been written by or with the help of AI chatbots. She recognized the robotic mimicry, Friedman wrote on her blog, because she’d had ChatGPT spit out writing in her own style.
The fake Friedman books were also, The Guardian reported at the time, taken down the author’s social media posts about them captured attention.
“I’m sure [the titles were removed] in no small part due to my visibility and reputation in the writing and publishing community,” Friedman wrote on her blog. “What will authors with smaller profiles do when this happens to them?”
In both of these instances of apparently AI forgery, Amazon told The Guardian that it is committed to taking down fraudulent titles, but all the same, an enforcement policy enacted after the fact — and often after authors or media contacts the company — seems too little too late.
In the less than two months between these stories, it’s clear that the AI book scammers are wising up – and it’s impossible to say how much better their cons will get in the future.
More on AI “writing”: Strike Ends With Writers Guild Forcing Concessions on AI
