VentureBeat presents: AI Unleashed – An exclusive executive event for enterprise data leaders. Network and learn with industry peers. Learn More
Amazon may be facing down federal antitrust charges over its online marketplace and advertisements, but that isn’t stopping the tech giant from releasing new features for said marketplace and the third-party vendors who sell products and advertise through it.
Today, in a post on the social network X from Amazon CEO Andy Jassy (who took over from founder and longtime CEO Jeff Bezos back in 2021), Amazon debuted a new generative AI feature that allows vendors to upload photos of their products to Amazon’s Ad Console service and then add AI-generated backgrounds.
How the new Amazon AI product image generator works
Amazon explained the feature in a blog post: “For example, an advertiser may have standalone images of their product against a white background, like a toaster. When that same toaster is placed in a lifestyle context—on a kitchen counter, next to a croissant—in a mobile Sponsored Brands ad, click-through rates can be 40% higher compared to ads with standard product images.”
In a promotional video showcasing the new feature, it appears that a user first navigates to the Ad Console, where they can create new advertisements for their products for sale through Amazon. Then, they enter at least three (or more) product serial numbers into an open text box under “Enter List” and click “Add.”
An exclusive invite-only evening of insights and networking, designed for senior enterprise executives overseeing data stacks and strategies.
From there, the products will appear in a new list to the user. Clicking a button marked “Go to creative” will bring them to a screen allowing them to “Customize Images” and enter in whatever prompt text they’d like for a background image to their product in a field marked “Image descriptions.” However, there is a 300 character limit.
Similarities to other image generators and additional features
The functionality works similar to other open-ended text-to-image generators such as OpenAI’s DALL-E 3 and Midjourney, in which users input descriptive text prompts, and the AI returns an image based on the descriptions.
In addition, Amazon’s promotional video also reveals that the new AI product image generator includes another feature: themes. Clicking “enhance with a theme” after generating a background will allow the user to further augment the image with additional props and objects that fit into different thematic categories, such as “Pumpkin spice,” which places realistic pumpkins all over the background to create an autumnal vibe for the user’s product image.
Amazon’s video shows a list of dozens of stock themes organized around loose settings and aesthetic families, from “cottage” to “forest” to “metallic” to “office.”
How it helps product sellers and advertisers
In its blog post, Amazon wrote that it built the new AI product image generator for the purpose of “enabling those that do not have in-house capabilities or agency support to more easily create brand-themed imagery.”
The feature, still in beta for now, comes on the heels of a similar one announced by Meta Platforms for its advertisers on Facebook and Instagram.
VentureBeat’s mission is to be a digital town square for technical decision-makers to gain knowledge about transformative enterprise technology and transact. Discover our Briefings.