Amazon Monetizes AI Shopping Engine—Again

Amazon built something powerful in private. Over three years, its AI shopping assistant helped more than 300 million customers find, consider, and buy products. It generated nearly US$12 billion in additional sales for the company in 2025 alone.

Now Amazon is selling access to that same engine to other retailers.

On May 27, 2026, AWS launched its Agentic Shopping Assistant (ASA). The product gives outside retailers the same conversational AI technology that powers Alexa for Shopping. It comes with a deployment window of roughly 60 days. What used to require years of engineering time is now a packaged service.

Why Amazon Is Licensing Its Retail AI

This is not an act of generosity. It is a calculated business move, and it follows a pattern.

Twenty years ago, Amazon built its own internal cloud infrastructure to run its retail operation. Then it packaged that infrastructure as Amazon Web Services and sold it to the world. AWS is now a US$128.7 billion annual business. The same logic applies here. If Amazon can sell access to its AI shopping engine and charge retailers for cloud usage, data processing, and support, it earns revenue on the technology twice over.

The routing is deliberate. The service runs through AWS, Amazon's cloud division, not through Amazon's retail arm. This matters to retailers who have long worried about sharing customer data with a competitor. Under this model, each retailer's data stays inside its own AWS account. Amazon sees the infrastructure usage, not the shopping behavior of individual customers.

Kate Spade Is First. Others Are Watching.

Kate Spade launched its AI Gift Concierge on April 13, 2026, making it the first production retail AI assistant built on Amazon Bedrock AgentCore. The tool is designed for a specific shopping challenge: gift buying. Most shoppers arrive with vague intent. The concierge turns that vague intent into a specific product recommendation through natural conversation.

The build took roughly two and a half months from contract to production.

Yang Lu, Chief Information and Digital Officer at Tapestry (Kate Spade's parent company), described the collaboration plainly: "AWS brought the recipe, but together we built the customization our consumers needed."

Tapestry's broader AI investment puts the Kate Spade launch in context. The group grew its digital revenue from US$600 million to US$2 billion over three years through AI-powered campaigns. The Gift Concierge is not a standalone experiment. It is the latest step in a company that has already committed to AI at the enterprise level.

The Conversion Data Is Significant

The business case for conversational shopping tools rests on one number. Conversational shopping sessions convert at 3.5 times the rate of traditional keyword search. Independent research from UK retailers puts the gap even higher, with conversational AI users converting at 200-300% higher rates than those using standard browse-and-filter interfaces.

Retail AI Personalization: From Clicks to Conversions

AI personalization in retail drives higher conversion rates. Learn how leading brands are implementing AI shopping assistants.

Mission MediaMission Media

A customer who can have a conversation about what they are looking for is far more likely to buy than one left to type keywords and scroll through product grids.

Brands using AI-powered search report conversion rates of 8-12%, compared to 2-3% for traditional site search. That gap explains why 84% of brands say conversational commerce is more strategically important now than it was a year ago, and why 82% expect it to be mainstream in their sector within two years.

The global conversational commerce market reached US$26.3 billion in 2025 and is on track to surpass US$30 billion this year.

What This Means for Retail Marketing Leaders in APAC

Retailers who do not build a branded conversational interface are not simply falling behind on a feature. They are handing the customer conversation to someone else.

AI platforms from ChatGPT, Google Gemini, and Microsoft Copilot are already completing real purchases on behalf of users. These platforms sit between the retailer and the customer. They curate, recommend, and redirect based on their own logic, not the retailer's brand voice or product strategy.

The AWS ASA offers an alternative: a branded interface that the retailer controls, built on product data and customer knowledge that third-party platforms cannot replicate.

This is not a question of whether to invest in conversational commerce. That decision has largely been made by market forces. The question now is where the conversation happens, and who controls it.