Key takeaways AI shopping agents threaten lucrative retail advertising businesses Amazon and Walmart dominate a $74.4 billion market Automated purchasing could shift control over product reco
Key takeaways
- AI shopping agents threaten lucrative retail advertising businesses
- Amazon and Walmart dominate a $74.4 billion market
- Automated purchasing could shift control over product recommendations
Amazon and Walmart have built enormous advertising businesses around a simple idea: companies pay to appear where customers are about to spend money. But what happens when customers stop browsing and let artificial intelligence make purchasing decisions instead?
That question now threatens a market worth an estimated $74.4 billion annually in the United States. According to Reuters' November 4 investigation, Amazon and Walmart together control approximately 88% of US retail media advertising spending.
The potential disruption comes from AI shopping agents capable of comparing products, evaluating prices and completing purchases without requiring customers to scroll through sponsored listings.
Why Amazon Makes Money Before Customers Buy Anything
When someone searches Amazon for headphones, the first products displayed are not necessarily the cheapest or highest-rated. Some are sponsored listings purchased by manufacturers hoping to attract buyers.
Those advertisements are valuable because they appear when shoppers already intend to make a purchase.
Retailers earn advertising revenue whether or not an individual click ultimately produces a sale, depending on the advertising arrangement.
Amazon has transformed this model into a substantial business. Advertising consistently represents approximately 10% of its total sales, although growth slowed to 11% in the third quarter of 2026.
The economics become more complicated when customers delegate purchasing decisions to software.
Imagine asking an AI assistant to buy the best wireless headphones under $150. Instead of browsing dozens of listings, the agent might compare specifications, availability and prices before selecting one product.
The customer could complete the purchase without ever seeing the advertisements that retailers previously sold.
Why Amazon Is Blocking AI Agents While Walmart Welcomes Them
Amazon and Walmart are approaching this potential disruption differently.
Amazon blocked Meta's Muse shopping agent from accessing its marketplace in September, arguing that the software operated without authorization and created customer-data concerns.
The company has also challenged Perplexity's shopping agent in court, although an earlier injunction was overturned on appeal.
Walmart, meanwhile, introduced infrastructure on October 21 allowing third-party shopping agents to interact with customer accounts under defined permissions.
The contrasting strategies reflect a broader challenge surrounding agentic payments: retailers must decide how much control to give external software while protecting customer relationships.
Morgan Stanley estimates that AI agents could handle 10%–20% of US e-commerce spending by 2030.
However, that projection does not mean advertising revenue will decline by an equivalent percentage.
Could Advertising Become Invisible to Consumers?
The industry's likely response may be to change how advertisements reach buyers rather than eliminate advertising altogether.
Amazon has already incorporated advertising into its own AI shopping tools, while Walmart is experimenting with sponsored content inside Sparky, its shopping assistant.
Brands may increasingly pay for product information to be considered by AI systems, creating new forms of commercial influence.
Companies such as Procter & Gamble, Nestlé and L'Oréal are already adapting product descriptions for AI-driven discovery.
That transition introduces new questions about transparency, particularly when autonomous purchasing involves financial decisions made with limited human intervention.
A similar challenge appears in AI financial agents, where permissions and accountability can matter more than the sophistication of the software.