Amazon Shoppers Embrace AI While Brands Go Agentic

It’s not just hype. Even reporters such as Ezra Klein are theorizing that artificial general intelligence (AGI) will be achieved within two years.

That has massive implications for how we shop online. This is true on both sides: for shoppers AND brands.

Kashif Zafar, CEO of Agentic AI-powered Ads platform, Xnurta, notes that the evolution of marketplace AI represents a “seismic shift.”

“In my view, generative AI was just the beginning. Now AI agents are unlocking new possibilities for customers, brands, and the marketplaces themselves,” said Zafar.

AI’s Growing Adoption with Amazon and Walmart Shoppers

Consumers are rapidly adapting to AI-enhanced shopping experiences. But there are a few key segments where that adaptation is moving more quickly.

Amazon and Walmart shoppers are leading the charge. According to a recent Prosper Insights & Analytics survey, 29.3% of US adults aged 18+ regularly use generative AI such as ChatGPT, that number is much higher for the average Amazon and Walmart+ subscriber.

34.8% of Amazon Prime subscribers already use generative AI, and, for Walmart+ subscribers, that number is 34.9%.

That makes Amazon Prime subscribers 18.8% (and Walmart+ subscribers 19.1%) more likely to use generative AI than the typical US adult aged 18+.

These shoppers are embracing features like personalized product recommendations and intelligent search results that narrow choices based on individual needs.

For example, just over a year ago, Amazon announced Rufus, an AI-powered shopping assistant trained on Amazon’s product catalog to answer customer questions and give product recommendations.

In October 2024, it was estimated that Rufus answered ~274.3 million daily questions, or ~13.7% of total Amazon searches based on an analysis of an AWS blog that highlighted how much compute Amazon had to allocate for Rufus queries during Prime Day.

“You really see consumers slowly, but steadily increasing adoption of AI chatbots,” said Zafar. “It’s not unrealistic to expect the percentage of Rufus queries to hit 20-25% in the next few years, and that’s not AI search counting off-platform AI queries such as those on ChatGPT. The total number of AI queries could be much, much higher.”

Notably, nearly one-quarter of U.S. adults (and one-third of Walmart+ subscribers) now say they actually prefer an AI chat program over a human for online shopping help.

Amazon’s AI Investment: The Ad Hype Cycle

Amazon’s leadership has been vocal about weaving AI into every facet of its business, especially advertising.

According to Xnurta’s analysis, Amazon executives dramatically ramped up their mentions of “AI” in earnings calls throughout 2023 and 2024.

“The chart below illustrates Amazon’s AI maturity curve,” said Zafar. “What you see there is one of two possible scenarios. One possibility is that there are two separate AI hype cycles at play, starting with a ‘generative AI’ cycle that’s crashed down to earth only to be replaced with the beginning of an ‘agentic AI’ cycle. The second possibility is that there’s a single AI hype cycle curve which is starting to mature into actual productivity more broadly. The shape of this curve does very much match that of the Gartner’s Hype Cycle, possibly indicating that we’re in the second scenario.”

This “AI hype cycle” isn’t just talk; it mirrors real investments. As part of its Q4, 2024 earnings call, Amazon earmarked over $100M in AI investments in 2025. Unsurprisingly, shortly after its first mention of “agentic AI” in its Q4, 2024 earnings call, Amazon announced a new group within AWS focused specifically on agentic AI.

Amazon’s advertising division, now a tens-of-billions juggernaut, has evolved into a suite of AI-powered marketing tools. A prime example is the Amazon Marketing Cloud (AMC) – an analytics platform where advertisers can analyze campaigns using Amazon’s shopper data. At CES 2025, Amazon unveiled a new generative AI feature for AMC: a SQL query generator that lets advertisers simply describe an audience or insight they want, and the AI builds the needed query in seconds​.

The future is now.

The Agentic Shift: How Brands Are Automating Commerce

As Amazon and Walmart build AI into the shopping journey, brands are starting to dabble with AI agents themselves.

Instead of manual keyword bids and budget adjustments, an AI agent can analyze performance data and adjust thousands of campaign parameters 24/7. Programmatic advertising has evolved such that algorithms optimize bids and placements in real time, often outperforming human media buyers.

A platform like Xnurta’s AI Copilot, for instance, can execute millions of bidding decisions across Amazon and Walmart per day. In the case of a campaign with SunnyDaze Decor, Xnurta’s AI Copilot helped drive a 408% rise in ad-attributable sales, outdoing their historical record by 95%, and increasing overall ROAS by 27%. Xnurta’s AI Copilot executed 3.3 million campaign changes in two months, equivalent to the work of 156 ad professionals working full-time. That’s time savings of over 50,000 hours.

But How Do Brands Optimize for AI Search?

While brands use AI to drive better advertising results, there’s a larger strategic question brewing:

How can brands optimize for AI search?

If Rufus takes up a larger percentage of searches on the Amazon home page, that presents an entirely different optimization opportunity.

With the traditional search bar, you have a field of possible options that compete on price, reviews, and product photos. It’s keyword oriented, and ads represent a large percentage of the top of search.

Now, with AI search, those factors still matter, but comparison shopping will be less likely. Consumers will ask an AI agent for a recommendation, it’ll give some options, but the field of consideration will be narrower.

They’ll be even less narrow if consumers start using AI agents themselves to start making automated purchases. That’s more likely for lower consideration purchases, such as low-cost recurring items like cat litter.

As Amazon has started introducing ads within Rufus as well, there’s a delicate balance that will need to be struck. Amazon wants to drive relevant results and build trust with AI-wary consumers, but they also want to drive advertising revenue in a new and growing segment of search.

To do that, says Zafar, Amazon and Walmart will need to optimize for relevance: “If search is getting smarter, more contextually relevant, and better at understanding customer need, then advertising will need to be just as relevant. How do you do that? By getting even more data-driven and real-time,” said Zafar. “That’s the only way you’re going to be able to surface your product within AI search for a particular audience.”

Conclusion

Amazon’s and Walmart’s aggressive AI adoption shows that this technology is now table stakes for retail success.

Consumers are embracing the convenience of AI, from chat assistants that answer questions in seconds to smart search that feels like a personal shopper. Brands, in turn, are leaning on agentic AI to automate and amplify their commerce playbooks. The end result is a more responsive, intelligent marketplace.

The key will be balance: a balance between relevant results v. relevant ads, and between humans in the loop who pull the strings v. autonomous AI agents who do everything.