AI ebook generator Automateed released platform data on July 7, 2026, covering more than 77,000 books created by users in 216 countries and territories, offering one of the more granular public looks yet at what people actually produce when a machine handles the writing. The company, which turns a stated idea into a complete, publish-ready book, said the analysis drew on AI-powered categorization of over 63,000 ebook titles, according to Automateed.

The headline finding runs against a common assumption. Non-fiction, not fiction, dominates the platform's output. Self-Help & Personal Development ranked as the top category, accounting for 10.8 percent of categorized titles, according to Automateed. Health & Fitness followed at 9.6 percent, and Business & Entrepreneurship took third place at 7.2 percent. Fiction and novels, by contrast, made up just over 5 percent of categorized titles, according to the company.

"People assume AI writing tools are mostly used for fiction experiments. Our data shows the opposite - users are building practical non-fiction: guides on health, money, faith, and personal growth," said Stefan Mitrovic, founder of Automateed. "Our goal has always been to build the best AI ebook generator for self-publishers who want to go from idea to published book without a team behind them."

What the numbers show

The dataset spans 77,636 books, gathered through anonymized, aggregated platform data and geolocated by IP address, according to Automateed. Roughly 6 percent of books carried an unknown location because of that method's limits. Ebook titles were sorted into more than 40 topic categories using AI classification, a process the company said underpins every category figure in the release.

Format breakdowns point toward a platform built around the standard ebook rather than illustrated or interactive formats. Standard ebooks made up 90 percent of all output, according to the company. Novels represented 5.4 percent, illustrated storybooks 2.9 percent, and coloring books 1.7 percent. That distribution suggests most users approach the tool the way they might approach a word processor: as a way to produce a plain, sellable manuscript rather than a visually complex product.

Beyond the top three categories, the analysis found two subjects performing better than might be expected. Religion & Spirituality accounted for 6.1 percent of categorized titles, and Psychology & Mental Health reached 5.5 percent, according to Automateed. Both outranked Fiction, Finance, and Marketing as standalone categories, a detail the company highlighted as evidence that faith and mental wellbeing content occupies more shelf space in AI-assisted publishing than commentators focused on business or fiction use cases might expect.

Geography of production

Creation activity spread across 216 countries and territories, though it concentrated heavily in a handful of markets. The United States led with 17.4 percent of all books created, according to the company. India followed at 10.2 percent, Brazil at 7.5 percent, Nigeria at 5.1 percent, and Indonesia at 4.1 percent. Combined, those four markets outside the United States accounted for more than a quarter of all books produced on the platform, a concentration Automateed framed as evidence that AI book tools are lowering publishing barriers fastest in emerging markets.

Regional specialization also emerged within the country-level data. Indonesia over-indexed in Education & Academic titles relative to its overall share of output, according to Automateed. Brazil showed a similar pattern in Cooking & Recipes. The United Kingdom, meanwhile, accounted for over a third of all Cybersecurity & IT ebooks created on the platform - a striking concentration given the UK's comparatively modest share of total books produced.

Methodology and its limits

Automateed described its methodology as anonymized and aggregated platform data covering the full set of 77,636 books, with categorization applied through AI classification across more than 40 topic buckets. The company disclosed the roughly 6 percent unknown-location rate tied to IP-based geolocation, a standard limitation for any tool relying on that method rather than user-declared location data.

What the release does not disclose carries its own significance for anyone trying to interpret the figures. Automateed did not publish a breakdown of how many of the 77,636 books were actually published on Amazon KDP or another retail platform, as opposed to generated and left unpublished. It did not disclose average length, price, or revenue per title. Nor did the release address whether books were purchased by real readers in meaningful numbers, or how many titles saw zero sales after creation - a gap that leaves open the question of how much of this output reached an actual audience versus sitting dormant in a self-publishing queue.

The company's own description of its product notes that Automateed supports ebooks, novels, storybooks, coloring books, and online courses, and is used by creators in more than 200 countries, according to the company's own materials. That places the disclosed 77,000-plus book figure within the context of a platform whose total addressable output likely extends well beyond ebooks alone, since courses and other formats fall outside this particular dataset's ebook-focused categorization exercise.

Context: a crowded self-help shelf

The finding that Self-Help & Personal Development leads Automateed's category breakdown arrives against a backdrop already documented elsewhere in the AI publishing conversation. Amazon's self-help market drowns in AI-generated content, reporting on a study from content verification company Originality.ai published January 28, 2026, found that 77 percent of books in Amazon's Success subcategory - a division of the broader self-help genre - were likely written by AI. That analysis, covering 844 books published between August 31 and November 28, 2025, also found that human-written titles in the same category received nearly five times as many reviews on average as likely AI-written ones: 129 reviews compared with 26.

The two datasets are not directly comparable. Automateed's release describes what its own users choose to create across a self-reported platform, while the Originality.ai research examined published, publicly visible listings already live on Amazon's marketplace and used AI-detection software rather than platform-disclosed data. Automateed's category figures cover ebooks broadly, whereas the Originality.ai study focused specifically on Amazon's narrower Success subcategory. Even so, both point toward the same underlying phenomenon: self-help and personal-development content sits at the center of AI-assisted book production, whether measured by what platforms report generating or by what detection tools find already on shelves.

That earlier research also documented a pattern of prolific single-author output that bears on how a figure like Automateed's 77,636 books should be read. One author identified in the Originality.ai study, Noah Felix Bennett, published 74 books between May and October 2025. Another, Richard Trillion Mantey, reached a total catalog of 397 books as of early December 2025. If similar concentration exists within Automateed's user base - a detail the company's release does not address - a relatively small number of highly active accounts could account for a disproportionate share of the platform's total book count, a pattern the release neither confirms nor rules out.

Why this matters for marketers

For marketers and publishers tracking the intersection of generative AI and content markets, Automateed's release offers a rare data point from inside a tool that produces the content itself, rather than from a platform measuring content after the fact. Most public research on AI-generated books to date - including the Originality.ai study referenced above - has approached the question from the demand side, scanning what already appears on retail shelves and inferring likely AI origin through detection software. Automateed's disclosure instead offers a supply-side view: a direct accounting, from the tool itself, of what categories and formats its users request most.

That distinction matters because it removes a layer of inference. Detection-based studies carry inherent uncertainty, since AI-writing detectors can produce false positives and false negatives, particularly on text that blends human editing with AI drafting. A platform-reported breakdown, by contrast, reflects what the company's own classification system recorded, though it introduces a different limitation: the analysis relies entirely on Automateed's self-disclosed methodology, with no independent verification of the underlying 77,636-book dataset or the AI classification system used to sort titles into categories.

The geographic concentration data carries separate implications. If AI book tools are indeed lowering publishing barriers fastest in emerging markets, as the combined 26.9 percent share from India, Brazil, Nigeria, and Indonesia suggests, that trend intersects with broader questions about content quality and market saturation that PPC Land has tracked across adjacent AI content categories. Whether self-publishing platforms in these markets will face the same scrutiny over AI-generated volume that Amazon's Success subcategory faced is not yet clear, since regulatory and platform-level responses to AI book content have so far concentrated on U.S. and European marketplaces.

The absence of sales, revenue, or reader-engagement data in Automateed's release leaves a gap that mirrors the one identified in the earlier Amazon research: creation volume and market success are not the same measurement. A book generated is not necessarily a book sold, read, or valued by a paying audience. Until platforms disclose that second layer of data alongside creation statistics, questions about whether AI-assisted self-publishing is producing genuine reader value or simply expanding catalog volume will remain open.

Timeline

  • August 31 - November 28, 2025: Study period covering 844 books analyzed in Amazon's Success subcategory by Originality.ai
  • January 28, 2026: Originality.ai publishes findings that 77 percent of Amazon Success subcategory books were likely AI-written
  • July 7, 2026: Automateed releases platform data covering more than 77,000 books created across 216 countries and territories

Summary

Who: Automateed, an AI ebook generator founded by Stefan Mitrovic, released the platform data. The findings concern creators who have used the platform across more than 200 countries.

What: Automateed disclosed anonymized, aggregated data covering 77,636 books created on its platform, categorized into more than 40 topic buckets using AI classification. Self-Help & Personal Development ranked as the top category at 10.8 percent, followed by Health & Fitness at 9.6 percent and Business & Entrepreneurship at 7.2 percent. Standard ebooks made up 90 percent of output, and the United States, India, Brazil, Nigeria, and Indonesia together accounted for the largest share of books created by country.

When: Automateed published the findings on July 7, 2026.

Where: The analysis covers books created by users in 216 countries and territories, with geolocation determined through IP address and roughly 6 percent of books carrying an unknown location.

Why: The release offers a supply-side data point on AI-assisted book creation at a moment when demand-side research, including a January 2026 Originality.ai study of Amazon's Success subcategory, has already documented heavy AI presence in the self-help genre specifically. Together, the two datasets suggest self-help and personal-development content occupies a central position in AI-assisted publishing, while neither dataset yet answers whether that volume translates into books that readers actually buy, read, or value.