Amazon announced a new feature this week that has left a lot of people doing a double take. When you search for something on the Amazon shopping app, you will now see AI-generated product images, pictures of products that do not actually exist, appearing below your autocomplete suggestions.
The idea, according to Amazon’s blog post, is to help shoppers who have something in mind but cannot find the right words to describe it. Say you are looking for a shirt with a specific neckline but you do not know it is called a “cowl neck.” Or you want rattan furniture but have no idea what rattan is called. Amazon says these AI-generated visuals will bridge that gap, you see a few style variations, tap the one that looks right, and get directed to real search results that match.
On paper, that sounds like it could be useful. In practice, it raises some obvious questions.
Customers who do not read carefully may think they are being directed to a page where they could find that exact product, then be disappointed when it is not actually available. That is a real risk. Someone sees a dress they love, taps it, and lands on a page full of similar-but-not-identical options. That gap between expectation and reality is exactly the kind of friction that makes online shopping frustrating.
There is also a more basic question worth asking: Amazon already has hundreds of millions of real product photographs. Why generate fake ones to guide people toward real ones? The whole point of shopping on Amazon is that what you see is what you can buy. Inserting AI-invented images into that experience feels like a step in the wrong direction, however well-intentioned the UX thinking behind it might be.
This feature sits alongside a growing pile of AI experiments Amazon has been running on its platform, some genuinely helpful, some harder to explain. On the more useful end, Amazon already summarizes customer reviews via AI, so you do not have to read through hundreds of them to get a sense of a product’s pros and cons. That one makes obvious sense. Fewer clicks, faster decisions, real information.
More recently, Amazon rolled out a short audio product summary feature where AI voices describe a product’s highlights in a podcast-style format. That one landed with a more mixed reaction.
Other recent additions include AI-generated shoppable collages for fashion styles, Amazon Lens Live which scans products through your camera to find visual matches, and a Lock Screen visual search widget for iOS. Earlier this month, Amazon also replaced its Rufus AI chatbot with Alexa for Shopping, enabling natural language queries by voice and text directly in the search bar.
The broader picture is that Amazon is clearly pushing hard to make AI a visible, central part of the shopping experience. Some of it works well. The review summaries genuinely save time. The visual search tools have a logical use case. But showing people images of products that do not exist, inside a store built entirely around products that do, that one still needs a clearer explanation.
Amazon has not shared data on whether features like this actually improve conversion rates or reduce search abandonment. Until they do, it is reasonable to wonder whether this is solving a real customer problem or just finding another surface to put AI on.
The feature is rolling out in the Amazon shopping app now.
