The AI Slop Gold Rush on Amazon: How One Journalist Found Her Own Fake Biography

In July 2026, The New York Times technology reporter Kashmir Hill published a revealing investigation into a fast-growing corner of Amazon’s marketplace: low-effort, AI-generated books that flood the platform with questionable content. What started as a personal surprise quickly turned into a window into a larger phenomenon of “AI slop” — cheaply produced books that mix public-domain facts with hallucinations, generic filler, and outright fabrications.

A Biography That Wasn’t Hers

It began when Hill learned from friends that Amazon was selling a book titled The Biography of Kashmir Hill for $26.99 in hardcover. Curious, she bought it. The 90-page volume was supposedly written by someone named John Crane Miller. Inside, she found a strange mix: some accurate details pulled from public sources about her life and career, heavily padded with complete nonsense.

The book invented elaborate personal rituals (including a coffee-making process that actually belonged to her husband), offered horoscope-style insights, and made sweeping claims about her “contradictions.” It was filled with the telltale signs of generative AI — repetitive phrasing, awkward em dashes, and confident-sounding statements that didn’t hold up to fact-checking.

Even more striking: John Crane Miller had published ten biographies in a single week, targeting other journalists and public figures. His author page featured a generic stock photo and a vague bio describing him as a “seasoned biographer and cultural analyst.” When Hill left a public review on Amazon asking to speak with him, the books (including hers) quickly vanished from the store, though his author page remained.

Meet Bill Johns, the Retiree AI Author

Hill’s investigation led her to a more cooperative figure: Bill Johns, a 70-year-old retiree living on the Chesapeake Bay in Maryland. After retiring from cybersecurity work in 2024, Johns got bored and decided to experiment with ChatGPT. He spent about $20 on the tool for his first book, a 651-page history of hacking called Ghosts in the Machine.

He hasn’t stopped since. As of the article’s publication, Johns had published 445 books on Amazon across wildly varied topics — sports, eccentric geniuses, famous bridges, alcohol, the Chesapeake Bay, and more.

On every cover sits the same AI-generated portrait of Johns in a sharp business suit. He admitted he couldn’t be bothered to put on real clothes and take a photo. His goal is to hit Amazon’s current maximum of ten books per week.

How the Business Actually Works

Johns’ process is straightforward and low-effort:

He feeds ChatGPT links and source material as “anchors.”
He generates chapters incrementally to reduce hallucinations.
He does light editing and declares on the platform that AI was used for “the entire work, with extensive editing.”
Books are uploaded via Amazon’s Kindle Create tool and sold primarily as print-on-demand paperbacks.

Financially, it’s modest but steady “bar money,” as Johns cheerfully calls it. He sells a few hundred books per month on average, earning roughly $7 per copy. During the 2025 holiday season, he sold 821 books and cleared nearly $6,000 in profit.

He removes titles when they receive valid one-star reviews, but otherwise keeps the operation running.

The Bigger Picture on Amazon

Hill’s story is part of a much larger trend. According to research cited in the article, e-book publications on Amazon roughly tripled after ChatGPT’s release, reaching over 300,000 per month by the end of 2025 (up from around 100,000 in 2022). Nonfiction appears especially vulnerable to AI flooding.

Other examples include:

A biography of Dan Rather that the veteran journalist actually read and promoted on Facebook.
A hastily produced book about the late Charlie Kirk that briefly became an Amazon bestseller before being removed amid scathing reviews calling it “AI slop” and “a scam.”
Prolific output from other pseudonymous authors like Diane W. Gray, who published dozens of celebrity biographies in a short period.

Amazon’s official position is that AI-generated content is allowed as long as it doesn’t create a “poor customer experience.” The company does not require sellers to label books as AI-generated, despite pressure from groups like the Authors Guild. A spokesman noted that Amazon invests resources in enforcement but resists mandatory disclosure rules that could punish honest users while rewarding those who lie.

Is This Just Digital Kitsch — or a Problem?

Not everyone sees AI slop as purely negative. Some philosophers and economists argue it lowers barriers to creation and expands the market, comparing it to historical “kitsch” that was once dismissed but later appreciated. Others, including Authors Guild CEO Mary Rasenberger, view it as soulless “word prediction” that degrades the platform and harms legitimate authors.

For now, the economics remain attractive for people like Bill Johns: minimal upfront cost, automated production, and enough sales to fund hobbies. The real question is how long Amazon will tolerate (or how effectively it can police) this flood of low-quality content before it affects customer trust or triggers stronger policy changes.

Kashmir Hill’s piece offers a clear, personal look at how generative AI is reshaping not just creative work, but the entire ecosystem of online publishing — one hallucinated biography at a time.

The full article is available on The New York Times website for subscribers.

---