AI-Generated Books Flood Amazon and Bestseller Lists, Threatening Human Authors' Livelihoods

AI-Generated Books Dominate Self-Publishing Landscape

The $140bn publishing industry faces an existential crisis as AI-generated books flood online marketplaces at an unprecedented scale. A groundbreaking study conducted by computer scientists at Stony Brook University analyzed 14,419 self-published fiction ebooks on Amazon using Pangram, an AI detection tool considered among the most reliable of its kind2

2

. The findings reveal that 2,168 books contained more than 50 percent AI-generated content, with 970 novels being more than 90 percent AI-written2

2

. The researchers estimate that sales of books containing at least 25 percent AI-generated text totaled tens of millions of dollars over the past three years.

"In genre fiction, people are using A.I. to write these books, and readers are not able to tell," said Tuhin Chakrabarty, a professor of computer science at Stony Brook University who led the research2

2

. The study found that AI-generated fiction now commands a 20% share of total sales in 2026, with a third of new entrants in bestseller lists containing AI content5

5

. This dramatic shift represents a fundamental challenge to traditional publishing models that have operated in similar ways for decades if not centuries.

The Daggermouth Phenomenon Raises Quality Questions

The case of Daggermouth illustrates how AI authorship has evolved beyond easily dismissable "AI slop" into commercially successful products. The self-published novel spent months on USA Today's bestseller list and ranks No. 1 in Amazon's science-fiction romance genre3

3

. Simon & Schuster paid seven figures to secure publishing rights to the book and its sequel3

3

. However, Pangram flagged the novel as 60 percent AI-written, and researchers discovered numerous "rare expressions" — identical five-word phrases appearing in other suspected AI books but almost never in human-written ones3

3

.

Author H. M. Wolfe denied using AI, stating through her lawyer that "The suggestion that I used generative AI to write Daggermouth is wholly untrue"3

3

. Yet the book contains sentences like "They collapsed together, a tangle of sweat-slicked limbs and racing hearts" that appear word-for-word in other self-published ebooks by different authors3

3

. This raises uncomfortable questions about whether AI-generated content can achieve commercial success while maintaining reader satisfaction.

Impact on Human Authors Becomes Devastating

The proliferation of AI-written books on Amazon is directly affecting human authors' ability to earn a living. Historical romance writer Jess Michaels, who has released 117 novels since 1999 and achieved 10 USA Today bestsellers, now faces such dismal sales that she might stop publishing altogether2

2

. "The glut of the A.I. work makes discoverability very difficult," Michaels explained. "You're talking about directly pirating my work, putting it into an engine so that you can produce a 'Jess Michaels-type book,' and then glutting the market with it. I cannot tell you how soul-crushing it is. It essentially erases actual authors"2

2

.

Research confirms these concerns, showing that AI-generated books have "cheapened the efforts of genuine authors," cutting what the average human-written book earns in most genres and pushing human titles down the bestseller ranks5

5

. The ethical implications for writers extend beyond immediate author earnings to questions about the long-term viability of writing as a profession.

Publishers Scramble to Address AI in the Publishing Industry

Traditional publishers find themselves caught between embracing technological efficiency and maintaining editorial integrity. "Every week we see a book that we suspect involves AI," says one senior book executive. "The grey area between generic writing and AI writing is difficult to judge. We don't know how much we unknowingly publish"1

1

. In March, Hachette was forced to pull Shy Girl, a horror novel by Mia Ballard, from shelves following allegations it had been partly written by AI1

1

.

Author Steven Rosenbaum's experience highlights the dangers even when AI tools are used with good intentions. After spending three years researching and writing The Future of Truth, he discovered his book contained fabricated quotations created when AI tools transformed summaries into what appeared to be direct quotes1

1

. "The danger isn't that AI writes bad prose," Rosenbaum observed. "The danger is that it writes persuasive prose attached to facts that never existed"1

1

.

Publishers are now inserting AI clauses into contracts. Pan Macmillan requires authors to guarantee their work is entirely human-authored to protect copyright1

1

. However, Madeline McIntosh, co-founder and CEO of Authors Equity, notes the complexity: "Unless what you are saying is 'you may only use pen and paper', the fact is you are touching AI -- and AI is touching your work"1

1

.

Amazon's Limited Response to AI-Written Books on Amazon

Amazon doesn't prohibit the sale of AI-generated books on its site. The company asks self-published authors to disclose AI use to Amazon but doesn't require them to flag it for readers2

2

. Amazon limits authors to uploading 10 books per week, but this still allows AI-assisted works at a scale impossible for human authors to match2

2

.

Amazon spokesman Josh Pflug stated the company has "clear content guidelines" and takes measures to prevent content that violates those guidelines2

2

. However, the company disputes research findings about title growth. A National Bureau of Economic Research study found that 53 percent of books published on Kindle in 2025 contained AI-generated text, compared to 23 percent in 20232

2

. The study also showed that the average rating of new books on Amazon has dropped as AI use has spread2

2

.

Detection Challenges and Copyright Battles Ahead

Spotting AI-generated content proves surprisingly difficult for average readers. When one reviewer searched Amazon specifically looking for AI-written books, they found that "at first glance, nothing looked unusual. All the books with AI-generated content featured professional-looking covers, convincing descriptions, reasonable price points, and even star ratings". The giveaways emerged only upon closer inspection: phantom authors with dozens of books published in weeks, generic marketing copy with phrases like "delve into" and "comprehensive guide," and fabricated quotations from public figures4

4

.

AI detection tools like Pangram offer some help but aren't perfect. Substack recently launched an AI detection feature in partnership with Pangram, though the platform acknowledges the tool's limitations4

4

. Meanwhile, copyright concerns escalate. Last week, Anthropic agreed to a $1.5bn settlement with authors and publishers after using over 480,000 published works to train its large language models without payment or permission1

1

. Both studies could fuel ongoing copyright cases against OpenAI, Anthropic, Google, xAI and Meta, where publishers and writers argue that the surge in AI-generated content has flooded the market and diluted their sales2

2

.