Quick Facts
- Amazon’s AWS launched the Agentic Shopping Assistant on May 27, using technology that drove nearly $12 billion in incremental revenue for Amazon last year
- Retailers can deploy AI shopping assistants in roughly 60 days, with conversational sessions achieving conversion rates 3.5 times higher than keyword searches
- Kate Spade parent company Tapestry launched an AI gift concierge using the service, with additional retailers currently testing the platform
Amazon Web Services announced the AWS Agentic Shopping Assistant on May 27, opening its internal AI shopping technology to retailers. The service builds on the same foundation as Alexa for Shopping, which served more than 300 million customers last year and generated nearly $12 billion in incremental sales for Amazon.
The new platform allows retailers to deploy AI shopping assistants in approximately 60 days. AWS claims conversational shopping sessions achieve conversion rates 3.5 times higher than traditional keyword searches.
“We are excited about the possibilities agentic commerce can bring to our customers,” said Yang Lu, chief information and digital officer at Tapestry. “AWS brought the recipe, but together we built the customization our consumers needed.”
Kate Spade launched an AI gift concierge in April using the technology. The assistant engages shoppers in conversation about occasions, recipients, and style preferences before recommending products. The implementation used Anthropic’s Haiku 4.5 model through Amazon Bedrock and required roughly 2.5 months of testing.
Each deployment is customized to match the retailer’s specific catalog, customer base, shopping environment, and brand voice. Retailers maintain control of their customer data, product catalogs, and business rules.
CEO Andy Jassy called the opportunity in agentic commerce “very good for customers” and said he expects it to benefit Amazon as well. “We have never seen a technology grow as rapidly as AI,” Jassy said.
The move follows Amazon’s strategy of commercializing internal technology, similar to AWS’s launch two decades ago and later offerings like cashier-less checkout and warehousing services. In return, Amazon gains access to first-party data from retailers using its technology.
The broader AI commerce market faces challenges. OpenAI shuttered its Instant Checkout initiative in March due to poor performance. Emily Pfeiffer, principal analyst at Forrester, said “nobody has figured it out, but everyone has FOMO. Everyone is prematurely rushing to market.”
McKinsey projects a $3-5 trillion global market opportunity by 2030. Competitors include OpenAI, Google, and Perplexity, which offer shopping research tools, while retailers like Walmart, Target, Etsy, Gap, and eBay pursue multi-platform approaches.
Amazon has positioned itself against open protocols by blocking OpenAI’s crawlers through robots.txt updates, removing 600 million products from ChatGPT’s shopping results. The company argues retailers should own the AI shopping experience rather than rely on external platforms.
Read more: Amazon opens its AI shopping tech for business with retailers

