On this episode of Self-Publishing with ALLi, Dan Holloway reports on a revealing piece by Rakuten Kobo CEO Michael Tamblyn, who explains why Kobo rejected 45 percent of self-published submissions in 2025 — most of them suspected AI-generated — and frames the decision as a “book community first” choice over a “readers first” approach. Dan also returns to the Commonwealth Short Story Prize controversy, where organizers have taken a strikingly different and very human approach to AI detection: gathering notes, drafts, and timestamped evidence from authors rather than relying on AI detection tools. He closes with news of a BISG and BookNet Canada survey on AI in publishing that indie authors in the US and Canada are encouraged to take part in.
Listen to the Podcast: Rakuten Kobo Took a ‘Book Community First' Approach to AI
Show Notes
Book Industry Study Group Survey
Sponsor
Self-Publishing News is proudly sponsored by PublishMe—helping indie authors succeed globally with expert translation, tailored marketing, and publishing support. From first draft to international launch, PublishMe ensures your book reaches readers everywhere. Visit publishme.me.
About the Host
Dan Holloway is a novelist, poet, and spoken word artist. He is the MC of the performance arts show The New Libertines, He competed at the National Poetry Slam final at the Royal Albert Hall. His latest collection, The Transparency of Sutures, is available on Kindle.
Thoughts or further questions on this post or any self-publishing issue?
If you’re an ALLi member, head over to the SelfPubConnect forum for support from our experienced community of indie authors, advisors, and our own ALLi team. Simply create an account (if you haven’t already) to request to join the forum and get going.
Non-members looking for more information can search our extensive archive of blog posts and podcast episodes packed with tips and advice at ALLi's Self-Publishing Advice Center.
Read the Transcript
Dan Holloway: Hello and welcome to another week's Self-Publishing News. I start with a story that illustrates the fact that the real news here in the UK is that we are no longer in a heatwave, and we are all wondering what on earth we did and said while our brains were fried. A huge thanks to Nate Hoffelder, who I often have occasion to thank, for pointing out one of my extreme brain lapses. In last Saturday's column, I referred to the CEO of Rakuten Kobo as Russ Tamblyn — who is, of course, the well-known actor. I should have said, and quite intended to say, Michael Tamblyn. With that introduction out of the way, let's have a look at what Michael — not Russ — Tamblyn had to say.
Kobo Rejected 45 Percent of AI-Generated Self-Published Submissions in 2025
This was a really interesting interview and piece that Tamblyn wrote for Publishing Perspectives, looking at Kobo's response to AI-generated self-published ebooks in 2025. The proliferation of AI-generated titles in 2025 has already made the news this year — starting with the announcement that four million titles were published in the US in 2025, up from three million the previous year, with the widely made assertion that AI-generated content drove much of that increase.
Tamblyn says that at Rakuten Kobo, they rejected 45 percent of submitted self-published books. As he puts it, ‘literally hundreds of thousands' — and most of those rejections were because, as he says, ‘best we could tell the books were generated with AI.' There's a huge amount to unpack in that statement and the piece generally.
One thing to unpack is whether there was collateral damage — a question that always arises when large platforms take large-scale action. Of those hundreds of thousands of rejected titles, how many were actually genuine books? This is causing much deliberation among people having to handle a flood of submissions. Authors get very upset when they feel they've been swept up in well-intentioned actions at scale. We've seen this before with reviews: whenever Amazon cracks down on fake or paid-for reviews, genuine reviews inevitably get caught in the net too. For authors, this matters not just for visibility, but for eligibility with promotional companies like BookBub, which require a certain review count. So that's one dimension.
The other dimension is why Kobo took this approach, and how it contrasts with James Daunt's statement earlier this year. Daunt said Barnes & Noble has a no-AI policy, but made clear that if readers decided they really wanted AI-generated titles, he would consider stocking them. Tamblyn does something more nuanced: he spells out that you can't keep everyone happy — which should be obvious, but often isn't, because people made unhappy don't realize that making them happy would make someone else unhappy.
As Tamblyn describes it, Kobo had an internal debate between a ‘readers first' approach and a ‘book community first' approach. The readers first approach simply asks: what do our readers tell us they want? Let's give them that. That's essentially Daunt's version. The book community first approach, as Tamblyn describes it, holds that authors and publishers need to be able to make a living. He doesn't say he's placing all his eggs in the creative basket and ignoring readers — rather, it's about where you place the emphasis and where you place the burden of proof. For Kobo in 2025, the emphasis was squarely on ensuring human authors and creatives get paid for their work and that the integrity of that work and that livelihood is protected. It's a really interesting and nuanced piece.
Commonwealth Short Story Prize: A More Human Approach to AI Detection
That brings us to a story we've covered before: the Commonwealth Short Story Prize. The prize is about to announce its overall winner, and when it announced its regional winners, there was significant controversy — the Caribbean regional winner, ‘The Serpent in the Grove,' was called out by many people upon publication as raising AI concerns. This put the prize organizers in a very difficult position, and it put Granta — the well-known literary magazine that published the story as part of a collaboration — in a difficult position too.
As a result, Granta has announced it is pulling out of all what it calls external collaborations — agreements to publish competition winners where Granta has no editorial say and the editorial decisions rest with the competition organizers. The controversy has been damaging enough that they've decided not to take part in such arrangements in future.
What's really interesting is how the prize organizers have responded since the controversy. When it first broke, they noted that authors had to sign a statement declaring no AI use — and at that point they had little choice but to take that on faith. Since then, ahead of the final announcement, they have not turned to AI detection software. Instead, they've undertaken what they describe as a detailed due diligence process — speaking to all the authors involved and gathering evidence of what was actually involved in their literary process. Notes, drafts, timestamped earlier versions of the submitted stories: the kind of byproduct you'd naturally accumulate while writing a prize-winning piece.
It's obviously not a scalable approach for a platform receiving hundreds of thousands of submissions — but for a prize, it's rigorous and deeply human. It sidesteps the crude ‘this person uses em dashes, therefore AI' accusations and instead does something closer to a job interview: you ask someone to walk you through their experience, and you quickly discover who's faked their way onto the form, because under questioning the story evaporates. A very different approach from Kobo's, but both responses to the same underlying problem. The contrast between them shows the scale of what we're dealing with.
BISG and BookNet Canada Launch an AI Survey — Indies Welcome
Finally, for those of you in the US and Canada specifically — the Book Industry Study Group and BookNet Canada have launched a survey inviting people in the publishing industry to share their views on AI. It's very clear from the survey that self-published authors are warmly welcome to take part, and it would be really good if you did, because then there will be a segment specifically representing indie author opinions that can be broken out of the results and discussed meaningfully. I'll send Howard the link for that. Do go and take part if you're eligible, and I very much look forward to speaking to you again at the same time next week. Thank you very much.
