Self-Help & Personal Development
The largest categorized topic group, led by practical personal growth titles.
10.8%
Automateed data report
AI Self-Publishing Statistics: State of AI Self-Publishing 2026
A snapshot of what creators are building with AI: more than 77,742 book projects, global activity across 216 countries, and a clear signal that AI book creation is moving beyond a narrow creator niche.
This page expands on Automateed’s public PRWeb release about AI ebook creation activity across 216 countries, with additional platform context and breakdowns.
Data snapshot: July 2026Report reviewed: July 16, 2026
Book projects analyzed
77,742
Includes ebooks, novels, storybooks, and coloring books created on Automateed.
Books with location data
72,897
Aggregated by IP geolocation where country-level data was available.
Countries represented
216
AI self-publishing activity is already global, not only US-based.
Top format
90.0%
eBooks dominate AI book creation, far ahead of novels and visual-first books.
Book formats
eBooks are the default AI publishing format
In this snapshot, traditional ebooks make up nine out of every ten AI book projects. That suggests most creators are using AI first for practical knowledge products, lead magnets, guides, and commercial non-fiction before branching into visual or fiction formats.
eBooks
69,962 created
90.0%
Novels
4,227 created
5.4%
Storybooks
2,221 created
2.9%
Coloring Books
1,332 created
1.7%
Research use
Questions this AI publishing dataset can and cannot answer
The dataset can describe creation behavior inside Automateed during the stated snapshot: which supported format a project used, whether country-level session data was available and how activity was distributed across the recorded categories. It can help product teams and researchers understand which creation workflows people begin and where adoption is occurring.
It cannot establish the size of the entire publishing market, the number of unique professional authors or the commercial success of an individual format. One person can create several projects, a project can remain unfinished and a generated manuscript can be substantially revised before publication. Creation volume should not be presented as book sales, reader demand or verified author income.
Comparisons with retailer catalogues require separate evidence because Automateed does not observe every later distribution decision. A creator may export to Amazon KDP, sell directly, keep the project private or abandon it. Researchers citing these figures should preserve the snapshot date, population definition and geographic limitations instead of reducing the result to “AI published 77,000 books.”
Future editions can become more informative by reporting consistent time windows and clearly separated completion, export, publication and sales events where privacy and data quality allow. Until then, the responsible conclusion is narrower: tens of thousands of projects across a wide geographic footprint show that AI-assisted book creation is already used for many publishing intentions, with text-led ebook workflows representing the largest observed share.
The report is therefore best used for directional product and publishing questions. It can show which creation paths deserve closer study and where authors appear to be experimenting, but it should not be used to forecast an individual book's revenue or to claim that one niche is commercially validated. Those decisions require separate retailer, audience and sales evidence collected for the specific market.
When citing the data, link to this methodology and the dated public release, preserve the denominator used for each percentage and avoid combining project, country and topic samples as if they were identical populations. Clear denominators make later editions comparable and allow readers to distinguish a genuine trend from a change in measurement.
How to read the report
Creation volume shows behavior, not automatic publishing success
A created project records that someone used a book workflow; it does not prove that the manuscript was completed, edited, published or sold. That distinction is essential when interpreting the 90% ebook share. Ebooks are the broadest Automateed workflow and include practical guides, lead magnets and other text-led projects, so they naturally capture more creation activity than specialized visual formats.
Country totals describe where attributed creation sessions originated, not author nationality, book language or buyer location. Location was unavailable for a portion of the total dataset, which is why the geography base is smaller than the full book project count. A creator traveling or using network infrastructure outside their home country may also be attributed differently.
Topic shares come from the categorized ebook sample described in the public press release. A book can touch several ideas, but the analysis assigns a leading category for reporting. The figures should therefore be used to understand broad platform patterns, not to calculate the exact commercial size of a publishing niche.
These limits do not make the dataset less useful; they define the question it can answer. It is evidence about what creators chose to begin building with AI on Automateed during the snapshot period. Sales demand, read-through, review quality and long-term publishing outcomes require separate datasets.
Practical implications
What authors, researchers and publishing teams can use this data for
For authors and creator businesses
The format data supports a practical conclusion: text-led ebooks remain the easiest entry point for turning expertise into a structured digital product. That does not mean authors should copy the most common topic. A broad category such as self-help or health still needs a specific reader problem, credible source material and a distinct promise. Use the format writing library to compare book structures, then validate the audience before committing to a long draft.
The global distribution also changes the publishing checklist. Language, currency, payout availability and local examples should be considered early rather than treated as translation tasks after launch. Authors selling directly can review worldwide author payouts; authors exporting to retailers should verify each store's current territory and metadata rules separately.
For journalists and researchers
Cite the dataset as an Automateed platform snapshot, include the collection label and avoid generalizing the proportions to the entire self-publishing market. The strongest supported statements are that the analyzed platform recorded 77,742projects across the four listed formats, that 72,897 had country attribution spanning 216 countries, and that ebooks represented 90% of the format snapshot.
Link to this report for the expanded tables and to the PRWeb release for the public announcement. When comparing future editions, check the denominator and snapshot date rather than treating a larger cumulative total as growth over a fixed interval. We will update the page only when a new aggregate export has been reviewed and the methodology remains comparable.
Global adoption
AI self-publishing is not limited to one market
The top ten countries account for 43,675 creations in this country-level dataset, but the full footprint spans 216 countries. The United States leads, while India, Brazil, Nigeria, and the United Kingdom show strong demand from creators outside the traditional US publishing center.
| Rank | Country | Total | eBooks | Novels | Storybooks | Coloring | Share |
|---|---|---|---|---|---|---|---|
| 1 | USUnited States | 12,701 | 10,971 | 1,017 | 412 | 301 | 17.42% |
| 2 | INIndia | 7,398 | 6,800 | 292 | 206 | 100 | 10.15% |
| 3 | BRBrazil | 5,448 | 5,192 | 84 | 99 | 73 | 7.47% |
| 4 | NGNigeria | 3,746 | 3,472 | 206 | 59 | 9 | 5.14% |
| 5 | GBUnited Kingdom | 3,448 | 2,993 | 256 | 133 | 66 | 4.73% |
| 6 | IDIndonesia | 2,994 | 2,869 | 75 | 42 | 8 | 4.11% |
| 7 | FRFrance | 2,147 | 1,933 | 147 | 34 | 33 | 2.95% |
| 8 | DEGermany | 2,121 | 1,836 | 105 | 87 | 93 | 2.91% |
| 9 | ZASouth Africa | 2,035 | 1,825 | 136 | 50 | 24 | 2.79% |
| 10 | PKPakistan | 1,637 | 1,495 | 49 | 67 | 26 | 2.25% |
Topic demand
The strongest categories are practical, not experimental
Automateed’s public PRWeb release summarized AI-powered categorization across more than 63,000 ebook titles. The pattern was clear: creators are using AI to build practical non-fiction more than abstract fiction experiments.
Self-help, health, business, faith, and mental health titles led the categorized sample. That matters for authors and marketers because those categories often map to advice, frameworks, coaching material, lead magnets, and sellable expertise.
Health & Fitness
A strong signal that creators are turning expertise into wellness guides and programs.
9.6%
Business & Entrepreneurship
Business books remain a natural fit for AI-assisted outlines, frameworks, and lead magnets.
7.2%
Religion & Spirituality
Faith-led publishing outranked several traditional commercial categories.
6.1%
Psychology & Mental Health
Mind-focused books ranked above fiction, finance, and marketing in the categorized sample.
5.5%
Regional patterns
Different markets are using AI books differently
Indonesia over-indexed in Education & Academic titles.
Brazil showed notable demand for Cooking & Recipes.
The United Kingdom accounted for a large share of Cybersecurity & IT ebooks in the categorized data.
Creators are choosing utility first
The dominance of ebooks points to practical publishing: guides, lead magnets, business books, educational resources, and niche knowledge products.
The market is globally distributed
Only 17.42% of country-attributed creations came from the United States. AI publishing demand is highly international.
Visual-first formats are early
Storybooks and coloring books are growing categories, but still a much smaller share than text-led ebook creation.
Methodology
This report uses aggregated, anonymized Automateed platform analytics. The book type dataset covers 77,742 created book projects. The geography dataset covers 72,897 book creations where country-level IP geolocation was available.
The public PRWeb release used a related platform snapshot covering 77,000+ books and AI-powered categorization of 63,000+ ebook titles into more than 40 topic categories. This page keeps the same public direction and expands it with additional Automateed context.
Counts represent creation activity, not guaranteed published sales volume. Individual user data, book content, private account details, and personally identifiable information are not included.
This is a fixed July 2026 research snapshot, not a live counter. We publish a new version only after validating a fresh aggregate export, so figures from different collection periods are not silently mixed together. This report page was last reviewed on July 16, 2026.
Turn your expertise into an AI-assisted book
Automateed helps authors create a free preview, then upgrade into full books, covers, formatted exports, author websites, and publishing workflows.
Create a free preview