AI-generated fiction is winning a growing share of Amazon sales, not getting ignored as "slop."

AI-written novels aren’t just flooding Amazon’s self-published fiction shelves. They’re selling. A study of more than 14,000 genre titles published between 2023 and 2026 found that books with heavy AI content now represent a large and growing share of actual revenue, not just catalog noise.

The numbers tell two stories at once. The number of books generating any sales grew nearly 20 times over the period. Revenue grew only nine times. That gap is dilution. More titles are chasing the same pool of readers, and each book is earning less on average. But the AI-heavy titles are still capturing a meaningful slice of that shrinking pie, which means readers are buying them, whether they know it or not.

This isn’t a quality debate. It’s a market structure problem. If you can publish 50 titles in the time it used to take to write one, you can test more covers, more keywords, more niches. You can fail cheaper and iterate faster. Speed alone becomes a distribution advantage, even if each individual book performs worse than a hand-written equivalent. The study, detailed in the full paper, makes it clear that output velocity is now a competitive moat in certain markets.

For anyone running a content business, editorial operation, or platform with user-generated material, this is the pattern to watch. Volume can win even when quality per unit drops, as long as the cost to produce and test stays low enough. The defense isn’t to ignore AI tools. It’s to build something AI can’t replicate at scale: curation, voice, trust, or a brand that means something before someone clicks. That’s the kind of filter we build into the Omni Command Centre, so teams can use speed without losing the signal that actually matters. If your business depends on content that people choose, you need a moat that isn’t just faster typing.