How Amazon Uses AI to Create 12 Million Product Listings
How Amazon uses AI to write product listings: in 2025, independent sellers created more than 12 million sales-ready listings with Amazon's generative AI tools. A seller types a few words, uploads one photo, or pastes a link to their own website, and AI drafts the title, bullet points, and description. The seller reviews it and decides what goes live.
Want your team to find work like this in your own company? See our AI Workshop for organizations.
Key numbers
12 million+ sales-ready listings created by independent sellers with generative AI in 2025 (Amazon, Apr 2026)
900,000+ sellers have used Amazon's AI listing tools (TechCrunch, May 8, 2025)
About 90% of the time, sellers accept AI-generated content with little or no edits (Amazon, May 2025)
60%+ of sales in Amazon's store come from independent sellers (Amazon, Apr 2026)
The old way: sellers wrote every field
A product without a listing cannot sell. A listing with missing details is hard to find and hard to buy from.
Each seller had to fill in titles, bullets, descriptions, and attributes one product at a time. The form could require fields, but it could not write them.
Over 95% of independent sellers on Amazon also sell in other retail channels (Small Business Empowerment Report 2025). The product information already existed, but had to be rewritten in Amazon's format.
How Amazon uses AI
1. Draft a listing from a few words or one photo
At the end of 2023, Amazon let sellers describe a product briefly or upload an image, and AI generated an Amazon-style title and attributes.
2. Create a listing from the seller's own website
In March 2024, sellers could paste a product URL from their website, and AI turned that page into an Amazon listing (TechCrunch).
3. Create many listings from a spreadsheet
Sellers upload a sparse spreadsheet and AI builds multiple enriched listings at once.
4. Improve live listings with Enhance My Listing
In May 2025, Amazon launched a tool that suggests better titles, attributes, and descriptions for existing listings. Sellers can accept, edit, or decline each suggestion (Amazon).
Every step follows one rule: AI drafts, and the seller, who knows the product best, approves.
Results
900,000+ sellers have used the tools, and in 2025 alone they created 12 million+ listings.
Amazon says sellers who create listings with its generative AI tools see a 40% increase in overall listing quality (Amazon).
In 2025, Amazon earned USD 172.16 billion from third-party seller services and USD 68.64 billion from advertising (Amazon 2025 Form 10-K). Both depend on products having listings that customers can find.
A caveat: TechCrunch notes that accepting AI content without edits is not proof of accuracy. Some sellers may not review the output closely.
What executives can take from this
Look for high-volume work where every item is different, such as onboarding product data from suppliers in many formats.
Let people start from what they already have, such as photos, specs, or web pages, not a blank form.
Shift people from writing to reviewing, and measure whether they actually review.
More cases in this series: How Walmart uses AI to design apparel and How Rohlik uses an AI agent in supplier negotiations.
Sources
This article is based on public information. IM Impower was not involved in this project.


