Secondhand book sales are booming. Is it because of AI? Partly, yes — but not in the way most people assume. Price and sustainability are still the biggest reasons readers buy used, and AI has layered a newer motive on top: a growing number of buyers are shopping secondhand specifically to avoid AI-generated books flooding new listings.
Short answer: The secondhand book market is projected to grow from $26.96 billion in 2025 to $28.8 billion in 2026, and researchers now list AI-driven recommendations as a forecast growth driver. But the bigger AI story is defensive: readers and even AI companies are buying pre-2022 print books specifically because they're guaranteed free of AI-generated "slop."
I track AI's effect on publishing and content markets for this site every week. When I tested this claim against real market reports, not the vibes-based version you see on social media, the picture split into two trends that get lumped together. One is old and boring: people buy used books because they're cheaper and greener. The other is new and tied to AI: a documented flood of undisclosed AI-generated titles on Amazon is pushing some readers toward books they can trust weren't written by a model.
Why secondhand book sales are climbing
The market growth is real, and it's not small. The global second-hand books market is on track to grow from $26.96 billion in 2025 to $28.8 billion in 2026, a 6.8% jump. The Business Research Company's July 2026 report projects it will reach $37.27 billion by 2030, a 6.7% compound annual rate. That report names "adoption of AI-driven book recommendations" as one of five growth drivers it expects, alongside e-commerce expansion and circular-economy adoption.
That AI-recommendations driver is mostly algorithmic, not editorial. BookTok's For You feed uses machine-learning ranking to resurface backlist titles. A title going viral there often means the only affordable copies left are used, since publishers don't reliably restock a five-year-old paperback overnight. That's a real AI effect, but it's indirect.
The older, bigger reasons haven't gone anywhere. Per WordsRated's most recent analysis (updated March 27, 2026), 92% of buyers cite lower cost as their main reason for choosing used books, and 83% cite environmental concerns. More than half of US and UK readers say they now buy more secondhand books than new ones. None of that needed AI to happen. It's been building since long before generative models existed.
The AI angle nobody's market report captures yet
Here's the part that doesn't show up in a growth-driver bullet list: readers are increasingly treating a used, pre-2022 copy as proof a book is human-written. We covered this in our piece on how generative AI floods and dilutes the market for books. A July 2026 study of over 14,000 self-published titles found new releases had grown 38-fold since 2023, while sales grew only 7-fold. Books with heavy AI-detected text now take close to 40% of Amazon's Top-25 rank slots in some genres, and almost none of it is disclosed to shoppers.
That distrust isn't just a reader habit anymore — it's now a business model. 404 Media reported on July 21, 2026 that AI companies are buying up old print books in bulk through brokers like ISBNdb. Why? Pre-AI books are guaranteed free of AI-generated contamination. ISBNdb pitches it to clients bluntly: "The world's best AI training data is sitting on a shelf." It describes older books as "curated, peer-reviewed, domain-specific human knowledge" that scraped web text can't match. If AI labs pay a premium for books precisely because they predate AI slop, it's not a stretch that human readers are starting to value the same thing.
What you'll need to check this yourself
You don't need any special tools to check this yourself, just three things. First, pull up a used bookseller's site. ThriftBooks lists more than 13 million titles on its own homepage and is a good place to start, since its catalog leans toward the pre-2020s backlist that's least touched by the AI flood. Second, have a phone handy with a barcode or photo-scanning app installed, since that's how you'll compare prices in the steps below. Third, set aside about 15 minutes. Most of the signal — AI flooding, algorithm-driven demand, resale pricing — shows up fast once you know what to look for.
Step-by-step: telling AI-driven demand from ordinary secondhand demand
1. Check whether the book's genre is AI-flooding territory
Fiction genres sold heavily through Kindle Unlimited — romance, fantasy, and thriller especially — are where AI-detected text has grown fastest, per the study cited above. If you're shopping in one of those categories, treat a suspiciously prolific "new" author with more skepticism than you would in, say, literary nonfiction, where AI-flagged volume is still comparatively low.
2. Read the free sample before comparing prices
Amazon's "Look Inside" on a new listing takes under two minutes and is the fastest way to catch AI's telltale signs — flat sentence rhythm, generic description, no specific detail a human author would need firsthand knowledge to include. Our guide to spotting AI writing covers the exact patterns to look for.
3. Compare the new listing's price against a used copy
Open the same title on a used marketplace. If the "new" listing is a recent, unranked self-published entry with an inflated price and no reviews, and a used copy of an established edition is available for less with a real publication history, that price gap is doing double duty — it's cheaper and it's a known-human-written text.
4. Try an AI-powered resale app on a book you already own
Scan a book's barcode in BookScouter to compare buyback offers across roughly 30 vendors. Or try a photo-based tool like Underpriced AI, which uses image recognition to identify an item and price it against real sold listings. Its free tier gives you three scans with no card required, and paid scan packs start at $4. Apps like this are part of why supply into the used market has gotten easier. Reselling a book used to mean guessing at a fair price. Now it takes one photo.
5. Watch what's spiking on BookTok versus what's actually new
If a title trending on BookTok is five or more years old, its algorithm-driven demand is a used-market story almost by definition — publishers rarely rush a backlist reprint fast enough to meet a viral spike, so secondhand copies absorb most of the demand first.
Example prompts you can copy
These work with any general-purpose AI assistant to speed up the checks above:
“ I'm looking at a self-published [genre] book on Amazon with [X] reviews published in [month/year]. List the specific stylistic patterns typical of AI-generated fiction, then tell me what to look for in the free sample to check for them. “
“ Compare buying a new self-published copy of a book at $[price] against a used copy of an earlier edition at $[price]. Given the AI-generated book flood on Amazon, what questions should I ask before choosing the new one? “
“ I want to resell [book title, condition]. Walk me through comparing buyback offers from BookScouter against a photo-pricing app, and tell me which is faster for a single book versus a bulk lot of 20+. “
These work because they point the assistant at a specific, checkable decision instead of asking it to explain the AI book problem in general, which tends to produce a vague answer.
Common mistakes to avoid
In my testing, the most common mistake is assuming every used-book purchase is an "anti-AI" decision. For most buyers it's still just price and sustainability, and reading AI motive into a $4 paperback purchase overstates the trend. The second mistake is trusting star ratings as a quality signal on new self-published listings. A book can rack up early five-star reviews before enough buyers finish it to notice thin, AI-typical prose. Third, people assume a used copy is automatically safe from AI content. That's only true for books published before generative AI became common in self-publishing, roughly 2022 onward — a used copy of a 2024 self-published title carries the same risk as a new one. Fourth, don't assume a barcode-scanning app values every book fairly. Older or niche titles often price better through a photo-based comps tool than a straight ISBN lookup, since barcode tools lean on buyback offers rather than real recent sold prices.
How AI shows up in the secondhand book boom
| AI-related factor | How it shows up | Effect on secondhand demand |
|---|---|---|
| AI-generated books flooding new listings | 38-fold growth in self-published releases since 2023, ~40% of Top-25 slots in some genres | Pushes trust-sensitive buyers toward known human-written used copies |
| AI companies buying training data | Brokers like ISBNdb sourcing pre-2022 books because they predate AI slop | Competes with consumer buyers for the same "clean" backlist inventory |
| BookTok/algorithmic recommendation | Machine-learning ranking resurfaces older, often out-of-print titles | New-copy supply lags viral demand, so used copies absorb the spike |
| AI-powered resale/pricing apps | Photo and barcode scanners price a book in seconds across dozens of vendors | Lowers the friction to sell, increasing available secondhand supply |
| Traditional price/sustainability motives | 92% cite cost, 83% cite environmental reasons (WordsRated) | Still the largest share of the boom — AI is additive, not the main driver |
Tools that make this easier
If you want to check whether a book is likely AI-generated before you buy it new, our guide to spotting AI writing is the fastest starting point. Our AI tool reviews hub covers the detectors worth trying as a second opinion. If you're a student weighing used textbooks against new AI-summarized study aids, our picks for AI tools for students show where AI genuinely helps versus where a used physical copy still wins. Writers curious about using AI honestly, instead of flooding a category, should start with best AI writing tools, which separates drafting help from wholesale AI-generated output. And if reselling books at volume interests you as a side business, best AI tool for small business is a good next stop for inventory and pricing tools that scale past a single scan. For the wider pattern of AI reshaping content markets beyond books, our AI news hub tracks it as it develops.
My take
Secondhand book sales are booming mostly for the reasons they've always boomed — used books are cheaper and greener, full stop. But AI has added a real, measurable second layer on top: a documented flood of undisclosed AI-generated titles has made "human-written" into a selling point again, and that's genuinely new. I don't think that's the majority of the growth yet, but it's the part worth watching, because the same AI companies buying old books to avoid training on AI slop are chasing exactly the same shelf space regular readers are.
Frequently Asked Questions
Is the secondhand book boom actually caused by AI, or is it something else?
Mostly something else — price (92% of buyers) and sustainability (83%) remain the top reasons people buy used, per WordsRated. AI is a newer, additive factor: a documented flood of undisclosed AI-generated books on Amazon is pushing some readers toward used copies they can trust are human-written.
How much has AI actually flooded new book listings?
A July 2026 study of over 14,000 self-published titles found releases grew 38-fold since 2023 while unit sales grew only 7-fold, and books with heavy AI-detected text now take close to 40% of Amazon's Top-25 rank slots in some genres, per our full breakdown of the AI book flood.
Are AI apps helping people sell used books, too?
Yes. Photo-based pricing apps like Underpriced AI use image recognition to price an item against real sold listings (free trial of 3 scans, packs from $4, or $5/month for 20 scans), and barcode tools like BookScouter compare buyback offers across roughly 30 vendors. Both lower the effort needed to resell, which adds supply to the secondhand market.
Is a used book guaranteed to be free of AI-generated text?
No — only if it predates roughly 2022, before generative AI became common in self-publishing. A used copy of a book self-published in 2023 or later carries the same AI-content risk as buying it new.
Will the secondhand book market keep growing?
Market forecasts say yes: The Business Research Company projects the second-hand books market growing from $28.8 billion in 2026 to $37.27 billion by 2030. Whether AI's share of that growth increases likely depends on whether Amazon starts showing AI-disclosure labels to shoppers instead of just collecting them at upload.