Artificial intelligence is moving from answering questions to making purchases. A new wave of AI shopping agents can browse product listings, compare prices, and even complete transactions on a person's behalf. That convenience could be a problem for the two biggest players in US retail media: Amazon and Walmart.
Together, the two retailers control an estimated 88% of the $74.4 billion US retail media advertising market, according to Reuters, citing data from Emarketer. Retail media is the business of selling ad space on a retailer's website and app—those “sponsored” product slots shoppers see while scrolling. It is a highly profitable business because it turns a retailer's existing digital shelf space into an auction marketplace. But AI agents don't scroll the way humans do, and that threatens the core of how retail media makes money.
Why AI agents are a threat to retail media
Retail media ads work best when a human is browsing. A shopper searching for “coffee maker” sees sponsored results at the top of the page, and advertisers bid for those slots in real-time auctions. The more eyeballs on those slots, the more advertisers are willing to pay.
When an AI agent does the shopping, those eyeballs disappear. The agent reads product data, compares options, and picks a winner—often without ever “seeing” a sponsored ad. That could undercut the auction-style pricing that makes retail media so lucrative.
Morgan Stanley, a global investment bank, has estimated that AI agents could account for 10% to 20% of US e-commerce spending by 2030. That is a big enough slice to matter to investors who have come to rely on retail media as a growth engine for Amazon and Walmart.
Two different strategies
The two retailers are responding in very different ways. According to Reuters, Amazon has blocked Meta's shopping agent and has challenged other agent tools in court. The company appears to be trying to limit third-party bots' access to its platform, protecting its ad business from disruption.
Walmart, by contrast, is embracing the trend. On October 21, the company rolled out tools to support “agentic commerce,” including linking an AI agent to a customer account with purchase approvals and spending limits. That approach lets Walmart keep a hand in the transaction even when a bot is doing the buying.
Amazon's caution is understandable given its recent ad performance. Reuters, citing LSEG data, reported that Amazon's ad revenue grew 11% year over year in the third quarter—the slowest pace in nearly four years. Advertising now makes up about 10% of Amazon's net sales, so any slowdown in that business has an outsized impact on profits.
What could change for advertisers
Ads may not disappear entirely, but they could change shape. If AI agents are reading product data rather than looking at images, retailers may need to provide more structured, machine-readable product information. That could shift the focus from flashy creative to clean data feeds.
Dan Marc of Footprints AI, cited by Reuters, expects more multiyear contracts and more discounting as agent-led shopping expands. Long-term deals can provide stable revenue, but they can also cap upside compared with a hot auction market. Discounting could squeeze margins, even if overall e-commerce sales keep growing.
Walmart's CFO, John David Rainey, said advertising (including Walmart Connect), membership, and marketplace drove 45% of quarterly operating income growth. That underscores how important retail media has become to the company's bottom line—and how much is at stake if ad prices soften.
What it means for investors
For investors, the rise of AI shopping agents is a double-edged sword. On one hand, agents could make shopping more efficient and drive more e-commerce volume. On the other, they could erode the high-margin ad revenue that has become a key profit driver for Amazon and Walmart.
If agents take over more searches, retail media auctions could lose pricing power. That would be a headwind for both companies, but especially for Amazon, which has been counting on advertising to offset thinner margins in its core retail business. The company's recent ad slowdown—even before agents become widespread—shows how sensitive this revenue stream can be.
Walmart's more open approach may help it adapt faster, but it also carries risk. By letting agents into its ecosystem, Walmart could see its ad inventory become less valuable if agents bypass sponsored slots. The company's move to link agents to customer accounts with purchase approvals suggests it is trying to keep some control over the shopping experience.
For everyday investors, the key takeaway is that retail media is not immune to technological disruption. AI agents are still early, but their potential to reshape how people shop—and how retailers make money—is worth watching. As Zalando's launch of an AI shopping agent shows, the trend is spreading beyond the US. And as recent market reactions to AI spending demonstrate, investors are paying close attention to how companies adapt.
The next few quarters will reveal whether Amazon's defensive stance or Walmart's open approach proves more effective. Either way, the $74.4 billion retail media market is entering a new phase—one where the shopper may not be human at all.
