Buying a KDP Business That Uses AI: 8 Due Diligence Questions

If a KDP business earns $60,000 a year and sells at 2.2 times profit, $12,000 of undisclosed AI-generated titles can justify cutting the price from $132,000 to $105,600. That worked example comes from Empire Flippers’ guide to buying a KDP business that uses AI, and it captures the guide’s argument: AI isn’t the problem, but how a seller used it, and whether they told Amazon, decides what a catalog is worth. For first-time buyers browsing Kindle publishing listings, it turns a vague worry into a checklist.

The Framework

Amazon allows AI-generated books but requires disclosure under its KDP content guidelines. AI-assisted content, where a person writes and AI edits or brainstorms, needn’t be disclosed. AI-generated text, images or translations must be, “even if you applied substantial edits afterwards.” The disclosure goes to Amazon, not readers, so buyers can’t check it from outside. Empire Flippers says KDP businesses have sold for an average of 2.2x annual net profit over three years on its marketplace, so the income needs to last a little over two years to pay back. The eight questions add to the usual KDP checks.

Catalog and Compliance

1. How much is AI-generated? Ask for a title-by-title breakdown of how text, covers, images and translations were made, with proof such as ghostwriter invoices or AI tool subscriptions. Focus on top sellers first.
2. Was it disclosed? This is the most important question, since undisclosed AI content becomes the buyer’s policy problem. Ask to see the disclosure answers for top titles on a screen share, and get the seller to confirm their accuracy in the purchase agreement. Fixes are repricing, republishing before closing, or an earn-out tied to titles staying live six months.
3. Do books break quality rules? Amazon’s content quality guide bars books not significantly different from others, too-short content, poor translations and excessive reuse. Ask for any removal notices and an export of KDP’s Quality Notifications Dashboard.
4. Are the books good? Read the top five to ten titles and the one- to three-star reviews. Be extra careful with health, money, legal and safety topics, and watch for invented credentials on pen names.

IP, Durability and Operations

5. What IP are you buying? The US Copyright Office’s January 2025 report said AI output is protected only where a human made enough creative choices; prompts alone aren’t enough. Mostly raw AI text may have little protection, shifting value to listings, reviews and the account.
6. How easy is it to copy? Search the top titles’ keywords for recent lookalikes, then look for moats: hundreds of reviews, followed series, an owned email list, print and audio editions.
7. Does revenue hold up without new releases? Ask for revenue by title by month, the share from books older than 12 months, and what happened when the seller stopped publishing.
8. Can you run the workflow? Get SOPs, prompts, templates and tool accounts transferred, and confirm someone reads each book before it goes live.

Where It Doesn’t Apply

The guide is written by a marketplace that sells these businesses, and it ends by recommending buying through Empire Flippers, which says it vets listings and handles the account transfer. Its own advice still applies: vetting isn’t a substitute for doing your own due diligence. The copyright points also come with the author’s caveat that they aren’t a lawyer. For choosing where to buy, Stack Archive compared marketplaces in Flippa vs. Acquire.com: Which Platform to Use for Your First Micro-Acquisition.

Source: Empire Flippers

Content created with AI assistance and reviewed for accuracy.