Amazon Tests Generative AI Capability With Customers
This week, Amazon took another step forward in bringing generative AI to its customers.
The e-commerce giant launched a consumer-facing generative AI-powered tool to answer customers’ questions about products.
More from Sourcing Journal
Consumers Say Selkie's AI-Aided Valentine's Drop Lacks Heart
Sayari Plans New Platform Launch, Geographic Growth Following $228 Million Investment
“Amazon has been using machine learning and AI for many years in virtually everything we do. We’re constantly inventing to help make customers’ lives better and easier, and are currently testing a new feature powered by generative AI to improve shopping on Amazon by helping customers get answers to commonly asked product questions,” an Amazon spokesperson told Sourcing Journal in a statement.
The feature has only been spotted in the company’s app; it has not yet made an appearance on its website.
The tool allows users to inquire about an item through the “Questions” feature on the product page. The search bar feature in this section of the app had previously aggregated the most relevant answers to the question posed through information from existing reviews, product descriptions and more.
With the existing feature, users could ask questions like “Would these pants be long enough for someone who is 5’8”?” or “What do customers think about the color of this product?” That would yield a list of responses for a user to click through to find the answer to their question.
But the new feature, which seems to be available only to some users as the company tests, allows shoppers to ask questions like, “Would this be a good dress for a black-tie winter wedding?” or “Would this jacket keep me warm at a cold, snowy football game in an outdoor stadium?”
The new AI-powered tool can also provide style recommendations, give advice on gifting, tell jokes and write persuasive arguments for purchasing a product.
But like any generative AI tool, it sometimes provides irrelevant, inconsistent or inaccurate responses. For instance, upon asking Amazon’s new AI tool to add a product to the cart, a user may receive an answer that says, “Sure, I’ve added the [product] to your cart. … Let me know if you need any other details or have changes to your cart.” In actuality, though, the product does not appear in the user’s cart.
Amazon did not make clear whether this feature would become permanent, nor has it disclosed whether it would be released on its mobile or PC websites.
The latest technology-backed element of the app is hardly Amazon’s first foray into AI. The company has previously enabled size recommendations based on past purchases and other users’ reviews, begun testing AI-generated product review summaries, launched tools to help sellers write listings and more.
On the enterprise side, the company has also introduced Bedrock, Amazon Web Services‘ generative AI offering.
With its new test, Amazon continues to play among an increasing number of companies working toward stronger customer experiences through AI.
According to recent Google Cloud research, 81 percent of retail decision makers feel urgency to adopt generative AI in their business, and 72 percent feel ready to deploy generative AI technology today.
