Talk to Me, Amazon Shopping App: How AI Could Sort Through All the Products You're Looking At

CNET

22-05-2025

Online shopping excursions can quickly get overwhelming as you plow through product pages to find what's really right for you. Now Amazon is trying out a new tool to help you figure out potential purchases. All you have to do is listen.
With Memorial Day deals in full swing, the e-commerce giant is making a trial run with short audio summaries for Amazon Shopping app users, delivered by "AI-powered shopping experts" that will analyze product details, customer reviews and other information they find on the web. "The feature makes product research fun and convenient — it's like having helpful friends discuss potential purchases to make your shopping easier," Rajiv Mehta, Amazon's vice president of search and conversational shopping, wrote in a blog post Wednesday.
Amazon is starting out the "hear the highlights" AI feature in the US for a "subset" of customers on a limited number products before rolling it out more broadly in the coming months. Those products include the Ninja Blender, OSEA Undaria Algae Body Oil, the Rain Showers Splash Pond Toddler Water Table, the SafeRest 100% Waterproof King Size Mattress Protector and the Shokz OpenRun Pro Open-Ear Bluetooth Bone Conduction Sport Headphones.
The new Amazon feature is another iteration in what has become an AI flood tide in online shopping. Google earlier this week introduced AI Mode, which is designed to suggest, select and pay for items for you, and also showed off a new AI service called Try On that depicts how clothes you see online will look on you. Walmart is working on its own AI shopping agent to perform tasks such as placing reorders and filling online shopping baskets based on user suggestions. And Shopify has launched Bluecore, which can answer shopper questions and recommend products. Klarna has an AI shopping assistant for comparison shopping.
CNET deals expert Adam Oram noted that Amazon's new audio summaries will have to prove their value for consumers.
"This feature may be helpful for some shoppers, as there's a lot of information on an Amazon product page and it can be difficult to parse all of it to find out what you need to know," he said. "But it's worth noting that there is a risk with it incorporating customer reviews, especially as these have been known to be gamed in the past. There's also the inherent risk of LLMs hallucinating, which may lead to inaccuracies if this is applied at scale, which is the opposite of what people making buying decisions need."
As always with generative AI tools, it's essential to not take results at face value and to do some double-checking.
"Using tools to help you gain a sense of what you're looking at quickly is generally a good idea, so long as you follow up with your own research to vet that the information is accurate," said Adrian Marlow, another member of the CNET Deals team.
Consumer skepticism could be at least a short-term hurdle for all AI e-commerce tools. A recent study commissioned by Akeneo, a product experience company, showed that only 45% of consumers have "some level of trust in AI-powered recommendations and chatbots to provide accurate product suggestions based on their interests and preferences." The research also found that only 38% of shoppers who've used AI chatbots reported being satisfied with the support they received.
Amazon, with net sales of nearly $638 billion last year, is incorporating AI into more and more sectors of its business. Rufus is a generative AI shopping assistant, Alexa Plus has been revamped and Audible will start using AI to narrate audiobooks.