Amazon is exploring a hybrid mode that would add AI-generated commentary directly above traditional search results for certain queries, according to The Information. The approach would let shoppers receive a conversational summary without being routed into Rufus as a standalone experience. Amanda Doerr, Amazon’s vice president of core shopping, confirmed the company is working out when to steer shoppers toward Rufus and when to keep them in conventional results, the outlet reported.
Amazon’s position is that traditional search and conversational AI serve different moments. A shopper searching for milk wants fast results and a clear path to checkout. Rufus fits better when the query is open-ended, such as researching hiking boots before a purchase decision, Doerr told The Information. Any shift would be gradual and vary by query type.
Amazon’s search bar anchors a $68.6 billion advertising business built largely on sponsored search and product ads. Inserting AI summaries would change how placements compete for attention. Google has already given AI prime real estate in its own search engine through AI Overviews and a dedicated AI mode. On Amazon, Rufus still largely sits apart.
Some users flagged this week that certain queries in the standard search bar returned a conversational summary alongside product recommendations, with an option to keep chatting. Amazon confirmed it tests features with subsets of customers, The Information reported.
Amazon is treating grocery as a distinct design, building what Doerr described as a basket-first experience: search for milk and the interface surfaces cereal and bananas alongside it.
Advertisement: Scroll to Continue
Beyond Rufus, Amazon has upgraded its core search engine to handle natural language queries and added thumbnail images to autocomplete suggestions, the company told The Information. Nearly 60% of U.S. customer searches begin with an autocomplete selection.
Rufus generated nearly $12 billion in incremental annualized sales in 2025, PYMNTS reported. Amazon also announced in March a partnership with feed management providers including Feedonomics to source product details directly from retailers rather than scraping, a step aimed at improving accuracy across its AI-powered features, according to PYMNTS.
