Rory Cellan-Jones calls Amazon AI-generated bio ‘utter tosh’

A growing number of authors are discovering that their hard-earned literary works are being replicated by artificial intelligence (AI) and sold on the world’s largest e-commerce platform, Amazon. Writer and journalist Rory Cellan-Jones recently had an unnerving experience when he stumbled upon a memoir on Amazon, a memoir he had co-authored, complete with a cover design created by an unfamiliar individual. His reaction was one of sheer bewilderment, as he pondered, “I thought: ‘This is strange – who’s writing a biography of me?'” according to the Observer.

Read: Amazon halts AI-generated books impersonating author Jane Friedman

Writing on his own blog, Cellan-Jones said that the unauthorised biography “Ruskin Park: Sylvia, Me and the BBC” was “utter tosh,” adding that after reading a passage to a group of authors during a festival, the audience found it “both hilarious and infuriating.”

Upon further investigation, it was revealed that the text within his memoir had seemingly been generated by AI, utilising tools such as ChatGPT, which allows individuals to produce extensive written content without meaningful work behind it. Amazon, upon realising the deceptive nature of these publications, swiftly removed the biography and other works attributed to the pseudonymous author. However, the issue remains far from resolved, with numerous AI-generated publications continuing to slip through the platform’s filters designed to identify substandard books.

On top of this, the veteran broadcaster received an email from Amazon recommending the fake work, rubbing “salt in the wound.” He raises the important question: what is the business model behind this venture, as the book is being sold for £10.89?

The rise of AI-generated literary rip-offs on Amazon

In a particularly egregious case, fifteen AI-generated books authored by someone using the pseudonym ‘Steven Walryn’ were published on Amazon, only to be taken down several months later. Similarly, renowned author Jane Friedman, celebrated for her writings on publishing, successfully compelled Amazon to remove five fraudulent titles falsely attributed to her, all created through AI.

According to the Guardian, Nicola Solomon, the CEO of the Society of Authors (SoA), commented, “Amazon is evidently grappling with substantial challenges posed by the influx of AI-generated products in its marketplace, and it appears to be lagging in its response.”

Amazon’s challenge of balancing AI-generated content with quality control

And it’s not the first time Amazon has grappled with AI-related concerns. In September, Amazon removed books that had the potential to be deadly. A number of mushroom foraging books appeared to have been written by AI, raising the concern that these books are not being properly vetted. In response, Amazon recently introduced a new policy, limiting self-published books to three-a-day, while calling for authors to be explicit in their use of AI.

Notably, this issue has not escaped the attention of some of the literary world’s most distinguished authors, including Margaret Atwood, Viet Thanh Nguyen, and Philip Pullman. These authors fear that their literary works are being exploited to train AI models without their consent, recognition, or compensation.

Read: Amazon’s AI-written mushroom foraging books could be ‘life or death’

In response to the uproar, an Amazon spokesperson asserted that the company is committed to enforcing content guidelines and eliminating books that violate them. The spokesperson clarified that while AI-generated content is permitted, it must adhere to Amazon’s content guidelines and not result in a subpar customer experience.

However, discontent among authors is not confined to Amazon alone. In the previous month, The Authors’ Guild, in collaboration with seventeen prominent authors, including Jonathan Franzen, John Grisham, George R.R. Martin, and Jodi Picoult, initiated a legal action against OpenAI in the Southern District of New York. The complaint alleges that OpenAI engaged in wholesale copying of the plaintiffs’ works without permission or compensation, subsequently using these copyrighted materials in the development of large language models.

Read: Authors’ pirated books used to train Generative AI

During the same period, authors Michael Chabon, David Henry Hwang, Rachel Louise Snyder, and Ayelet Waldman also brought forth a lawsuit, accusing OpenAI of deriving benefits and profits from the “unauthorized and unlawful utilization” of their copyrighted content.

As the battle between authors and AI-generated content continues to intensify, it raises profound questions about copyright, intellectual property, and the ethical boundaries of AI development. Authors, united in their pursuit of justice, are determined to protect their creative works in the face of technological advances that threaten to exploit their intellectual property.

How To Be Books has contacted Amazon and are awaiting a response.