If you’ve written a book with AI assistance, finding a publisher is harder than it should be. Most publishers don’t have a public AI policy. The ones that do are mostly bans — on AI-generated manuscripts, AI-edited queries, sometimes even AI-assisted brainstorming. Agencies are adding blanket restrictions to their submission guidelines faster than authors can keep track.
The result is a landscape where authors who used AI responsibly in their process have to guess whether a publisher will reject them on principle before reading a single page.
You shouldn’t have to guess. Here’s what to look for.
Check their stated policy.
Start with the submission guidelines, not the marketing page. A publisher that welcomes AI-assisted manuscripts will say so. If their guidelines are silent on AI, that’s not a green light — it’s ambiguity that can turn into a rejection or a cancelled deal after the fact.
Look for explicit language. Do they distinguish between AI-assisted and AI-generated? Do they require disclosure, and if so, is it informational or gatekeeping? There’s a difference between “tell us what tools you used” and “we’ll run your manuscript through detection software before we read it.”
If you can’t find a clear answer, ask before you submit. An email to the submissions address takes five minutes. A cancelled deal after months of work takes a lot more.
Ask about detection software.
Some publishers and agencies now run manuscripts through AI detection tools before acquisition. That should matter to you, because these tools don’t measure what publishers think they measure.
Detection software looks at statistical patterns in sentence structure and word frequency. It can’t determine whether a human directed the creative process. It flags non-native English speakers, neurodivergent writers, and authors who happen to write clean, structured prose. The false positive rate is high enough that the tools’ own developers discourage using them as definitive judgments of authorship.
If a publisher makes acquisitions decisions based on detection scores, they’re not evaluating your manuscript — they’re evaluating a probability guess about your writing process. That’s a red flag regardless of whether you used AI.
Understand the deal structure.
AI-friendly is meaningless if the deal is bad. The publishing industry has a spectrum of business models, and not all of them work in the author’s interest.
Vanity presses charge you to publish. They make money from your fees, not from selling your book. If a publisher asks for thousands upfront to print and distribute your manuscript, they’re a vendor, not a partner.
Hybrid publishers split the difference — some upfront cost, some revenue share. The quality varies widely. Some are legitimate operations with strong distribution. Others are vanity presses with better branding.
A traditional publishing partner invests alongside you. They cover editing, production, and distribution at their own expense and make money when the book sells. The structure aligns incentives: they only profit if your book reaches readers.
When evaluating a deal, look at the revenue split, the rights term, and reversion. What percentage do you keep? How long does the publisher hold your rights? What happens when the contract ends — do your rights revert automatically, or do you have to fight for them?
Look at what they actually provide.
Distribution is table stakes. Getting a book onto Amazon is something you can do yourself in an afternoon through KDP. The question is what else a publisher brings.
Editing — developmental, copy, and proofread. Cover art that’s professionally designed, not templated. Interior formatting for both print and digital. ISBN registration under the publisher’s imprint. A dedicated author site with a sales funnel. Social media management. Bookstore outreach to indie stores and chains. Critic and influencer submissions to drive reviews.
The gap between “we’ll put your book on Amazon” and “we’ll publish your book” is the entire production and marketing stack. If a publisher’s value proposition ends at distribution, you’re paying a revenue share for something you could do for free.
Ask about licensing and adaptation.
This is where most publishers stop and where the real value starts. If your book has IP potential — a world that could become a series, a story that could become a film or show — does the publisher help you get there?
Licensing packaging means pitch decks, world bibles, lore books — the materials production companies need to evaluate an adaptation. Licensing outreach means actively pursuing those deals, not waiting for someone to call.
Most publishers don’t do this. They stop at the bookshelf. If your book has legs beyond print, you want a partner who builds for that.
Where NX stands.
NX publishes AI-assisted books with zero upfront cost. Same full-stack deal for every author — editing through licensing. We don’t run detection software. We evaluate manuscripts on quality, not tooling.
Read our full AI manuscript policy or pitch us your book.