There was a time when publishing ten books on management implied at least one of three things: a long career, a long argument with an editor, or an alarming inability to stop writing. Today it may imply something simpler: an internet connection, a KDP account and an AI tool that never asks whether you have actually managed anyone.
That does not make self-publishing the villain. Kindle Direct Publishing did something genuinely important: it broke the old bottleneck in which a handful of publishers decided who was allowed onto the shelf. Amazon lets people self-publish eBooks, paperbacks and hardcovers without paying a traditional publisher, and gives them access to a global retail platform. For unknown writers, specialists and authors outside the old publishing centres, this was liberation.
Then generative AI arrived and discovered that liberation also scales.
The modern management-book production line can be magnificently efficient. Pick a subject — leadership, negotiation, agile transformation, emotional intelligence, strategic resilience, high-performance culture. Ask a model for a structure. Expand the chapters. Generate case-study-shaped anecdotes. Polish the prose. Produce an English version, even if English is not the operator’s working language. Add a cover with a mountain, chess piece, lighthouse or person in a suit walking toward an abstract sunrise. Upload. Repeat.
The miracle is not that a machine can help write a book. The miracle is how quickly a book can now help write a biography.
On a marketplace page, “author of 14 books on leadership” looks like evidence. It is a number, and numbers arrive wearing little uniforms of authority. Fourteen sounds more serious than one. Thirty sounds almost professorial. Nobody needs to know whether the author has run a team, hired anyone, fired anyone, made payroll, survived a failed product launch or sat through the peculiar spiritual crisis known as a quarterly budget review.
Amazon does verify identities in some circumstances and requires tax and payment information for KDP accounts. What its public onboarding rules do not require is a management degree, an executive career or proof that the person publishing “Advanced Corporate Strategy” has ever advanced a corporate strategy. That is not a criticism of Amazon so much as a description of self-publishing. A bookstore is not a professional licensing board.
The trouble begins when readers forget the distinction.
Generative AI has made the gap between presentation and provenance cheaper to manufacture. Amazon now explicitly requires KDP users to disclose AI-generated text, images and translations to the platform. It distinguishes that from AI-assisted work, where a human creates the content and uses AI to edit, refine or brainstorm. The rule itself is revealing: the publishing system has had to create a new administrative category for books that can be materially generated by software.
And the speed is no longer hypothetical. KDP’s current help pages say creators are limited to 10 new titles per book format each week, with the possibility of requesting an exception if they regularly create large numbers. A limit of ten books a week is a wonderful sentence to place beside the romantic image of the tortured author staring at one unfinished manuscript for seven years.
Amazon has tightened rules over time. In 2023, amid concern about AI-generated books, it introduced disclosure requirements and a lower creation limit. The present rules have evolved again. Meanwhile, the Authors Guild — an advocacy organisation for writers and therefore not a neutral observer — said in its 2025 annual report that AI-generated books on Kindle and other stores had “skyrocketed”. It described knockoffs, misleading titles and scam books that could be produced fast enough to catch the attention surrounding real releases.
None of this proves that a prolific self-published management author is a fraud. Some experienced executives write quickly. Some excellent authors have no formal degree. Some non-native English speakers produce better English than native speakers. Some use AI responsibly as an editor or research assistant. Traditional publishing, for its part, has never been a perfect machine for filtering wisdom from nonsense; it has merely been slower and more expensive nonsense when it failed.
But generative AI changes the economics of appearing qualified.
Previously, manufacturing a shelf of books required time, language competence, ghostwriters, editors or money. Now a large part of that apparatus can be rented by the prompt. The marginal cost of one more title collapses. Once the first template exists, “Leadership for Remote Teams” can acquire cousins: “Leadership for Hybrid Teams”, “Leadership for Resilient Teams”, “Leadership for Innovative Teams” and, after lunch, “Leadership for Leaders”. The bibliography grows faster than the biography.
This creates an odd reversal in business publishing. Books used to be downstream of expertise: you did the work, accumulated experience, formed an argument, then wrote the book. Increasingly, the book itself can be upstream of the expertise brand. Publish enough titles, build an author page, quote your own frameworks on LinkedIn, appear on podcasts, add “international author” to the bio — and the market may infer the career that the books once documented.
The English-language market is especially attractive because it is global and prestigious. AI-generated translation and drafting reduce the old language barrier dramatically. KDP’s own policy explicitly treats AI-generated translations as generated content. That means a person no longer has to possess strong English prose skills to produce an English-language manuscript. The software can supply the sentences; the cover can supply the gravitas; the metadata can supply the discoverability.
What it cannot automatically supply is experience.
That is the part readers now have to audit themselves. A management book should trigger questions that an author count cannot answer. What organisations has this person actually worked in? What decisions did they own? Are the case studies traceable? Are claims supported by evidence? Does the author distinguish research from personal opinion? Is the “framework” more than four familiar nouns arranged around a circle?
The great democratisation of publishing has created more voices, which is good. The great automation of writing has created more output, which is not the same thing.
The next phase will probably not be a war between “real authors” and machines. It will be a quieter sorting problem. Readers, employers and event organisers will need better signals for expertise than the existence of a paperback with a confident subtitle. Platforms will keep adjusting rules. Writers’ groups will keep pushing for disclosure. And somewhere, while everyone debates the future of authorship, a newly minted management guru will be finishing Book 27.
Book 28 is already in outline.