Amazon is rolling out AI-generated product images in its shopping app, and the move is raising eyebrows among retail observers who question why a massive e-commerce platform would show shoppers fake photos instead of real ones. Announced on Wednesday, the new feature displays synthesized images of products beneath a user's autocomplete suggestions when they type a search query. So if someone searches for something like a "blue gingham dress" without knowing the exact style they want, they'll see several AI-rendered dress variations—different sleeve lengths, hemlines, and patterns—pop up as clickable visual options to steer them toward more relevant listings. The use case Amazon describes is users who have a vague idea of what they want but lack the vocabulary to find it. The company gave examples like searching for "cowl neck" shirt styles or "rattan" furniture—terms a casual shopper might not think to use. Clicking on one of the generated images would take users to actual product listings that match that visual style, leveraging Amazon's existing visual search technology. The problem, as critics have already pointed out, is that the feature essentially fabricates products to bridge the gap between what shoppers imagine and what Amazon actually sells. Customers who skim past the fine print could easily mistake the AI mockups for real inventory, only to land on results that don't quite match the image they clicked. The feature joins a growing pile of AI tools Amazon has been stacking onto its retail site and app in recent years, with results ranging from genuinely useful to outright odd. On the practical side, the company already uses AI to summarize customer reviews, condensing thousands of opinions into digestible pros-and-cons snippets. It also launched an audio product summary feature last year, where AI-generated "experts" deliver podcast-style recaps of product highlights. Other recent additions include AI-generated "shoppable collages" designed to funnel shoppers toward curated fashion pages, and Amazon Lens Live, which lets users scan products in real time through their phone cameras. Whether fake product thumbnails will land somewhere on that spectrum—or simply leave users confused—remains to be seen.