The rapid development of artificial intelligence shopping agents could change one of the most profitable aspects of online retail: the ability of major retailers to influence what consumers discover and purchase. For companies such as Amazon and Walmart, digital advertising is no longer simply an additional source of revenue. It has become an important part of their business models, allowing brands to pay for visibility while retailers collect valuable information about consumer behaviour. The emergence of artificial intelligence tools that can independently research products, compare prices and complete purchases threatens to weaken this relationship.
The potential disruption is not necessarily about consumers abandoning online shopping. It is about changing where and how purchasing decisions are made. If customers increasingly rely on artificial intelligence assistants to identify products, they may have less reason to visit retailer websites or browse their applications. That could reduce the number of opportunities retailers have to display sponsored products and influence buying decisions.
Amazon and Walmart are approaching this development differently. Amazon has resisted certain third-party shopping agents, while Walmart has started creating infrastructure that allows external artificial intelligence agents to interact with customers’ accounts. Their contrasting strategies reflect a deeper disagreement about who should control the next generation of online shopping.
The advertising business is built around customer attention
Amazon and Walmart occupy an unusually powerful position in the American retail advertising market. Together, they account for an estimated 88% of a market valued at approximately $74.4 billion in 2026. Their influence comes from controlling both the retail environment and the information generated when consumers search for and purchase products.
Traditional online advertising generally depends on attracting attention. Retailers can display sponsored products when shoppers search for particular categories, recommend items based on previous purchases and promote brands through personalised advertisements. Companies are willing to pay for this access because advertising appears close to the point where consumers make purchasing decisions.
This model also provides retailers with an important advantage over many conventional advertising platforms. They can connect advertising exposure with actual sales data, giving brands a better understanding of whether promotional spending has resulted in purchases.
Artificial intelligence shopping agents could weaken this advantage. Instead of visiting several retailer websites, consumers may simply tell an assistant what they need, specify a budget and allow the software to compare available products. The agent could then identify an appropriate item and complete the purchase with limited involvement from the consumer.
In such a situation, the retailer may still supply the product, but it may no longer control the customer’s discovery process. That distinction could have significant consequences for advertising revenue.
The threat comes from losing control of product discovery
The conventional online shopping experience gives retailers several opportunities to influence consumers. A customer searching for a television, for example, may encounter sponsored listings, recommendations, discounts and competing products before making a purchase.
An artificial intelligence agent can potentially compress this process into a single recommendation. It may compare technical specifications, customer reviews, delivery costs and prices across several retailers without requiring the consumer to visit each website.
The implications for advertising are considerable. If the software selects a product based primarily on relevance, price and specifications, traditional sponsored placement may become less effective. Retailers could lose some of their ability to charge brands for prominent positioning within their own digital marketplaces.
There is also a question of data ownership. Retailers have traditionally collected information about browsing activity, searches, abandoned shopping carts and completed purchases. This information helps them develop targeted advertising products and understand consumer preferences.
When an external agent manages the interaction, some of that information may become concentrated within the artificial intelligence platform instead. The retailer could retain transaction data while losing visibility into the consumer’s broader decision-making process.
This is particularly important because access to customer information is one of the foundations of retail media advertising. Without detailed information about what consumers want and how they make decisions, retailers may find it harder to offer advertisers the same level of targeting and measurement.
Amazon and Walmart are making different strategic choices
Amazon’s response reflects its interest in protecting control over its retail environment. The company has blocked certain third-party agents, including Meta’s Muse, arguing that external services should respect the rules established by the websites they access.
The company has also been involved in a legal dispute concerning Perplexity’s Comet shopping agent. A court order that initially blocked the agent was overturned by an appeals court in August, although the dispute remains ongoing.
Amazon’s position is understandable from a commercial perspective. Its retail platform is not merely a place where products are sold. It is also an advertising network, a marketplace for third-party merchants and a source of extensive consumer information. Allowing external agents to control customer interactions could weaken several of these businesses simultaneously.
However, completely restricting external agents also carries risks. Consumers may increasingly expect artificial intelligence assistants to work across different retailers. If Amazon makes that process difficult, customers could shift their shopping activity towards platforms that provide greater flexibility.
Walmart appears to be pursuing a more accommodating strategy. Its infrastructure allows third-party artificial intelligence agents to connect with shoppers’ accounts, while consumers can approve purchases or establish spending limits.
This approach could help Walmart remain part of the purchasing process even when consumers begin their searches outside its own website. Rather than attempting to prevent the development of external agents, the company is positioning itself to participate in the emerging system.
Neither strategy is without uncertainty. Amazon risks appearing restrictive, while Walmart may have to share more control over customer relationships with external technology companies.
Retailers could lose advertising revenue without losing sales
An important distinction is that the growth of artificial intelligence shopping does not automatically mean that Amazon and Walmart will lose their retail businesses. Consumers may continue purchasing products from these companies even when an external assistant manages the search process.
The more immediate risk concerns the economics of advertising. Retailers could continue generating revenue from product sales while losing some of the advertising opportunities that currently make their platforms especially profitable.
Amazon’s advertising revenue grew by 11% year over year in the third quarter, its slowest rate of growth in nearly four years. The company has said that third-party agents have not yet had a meaningful effect on its advertising business. It would therefore be premature to attribute the slowdown directly to artificial intelligence shopping agents.
Nevertheless, the development creates a strategic problem. Retailers need to protect existing advertising relationships while preparing for a future in which customers may interact with products through conversational software rather than conventional search results.
Walmart’s advertising, membership and marketplace operations have also become important contributors to operating performance. Its finance chief indicated that these activities accounted for 45% of quarterly operating income growth. This makes the protection of retail media revenue particularly relevant to the company’s wider financial strategy.
Advertising may shift towards artificial intelligence systems
The next stage of retail advertising could involve a different relationship between brands, retailers and consumers. Instead of paying primarily for prominent product placement, companies may increasingly compete to ensure that their products are accurately represented and recommended by artificial intelligence systems.
This would place greater importance on product information. Specifications, pricing, availability, compatibility and reliable descriptions could influence whether an artificial intelligence assistant selects one product over another.
Manufacturers are already examining how their product information can be made more accessible to artificial intelligence systems. Companies such as Procter and Gamble, Nestle and L’Oreal are among those adapting their digital information strategies to changing methods of product discovery.
Retailers are also experimenting with advertising within their own artificial intelligence assistants. Walmart has been testing advertisements in Sparky, its shopping chatbot, while Amazon has incorporated advertising into its own artificial intelligence shopping services.
These developments suggest that retail media may not disappear. Instead, the commercial value could move from controlling a website’s search results towards influencing recommendations made by intelligent shopping systems.
There are still unresolved questions about transparency. Consumers will need to know whether an artificial intelligence assistant recommends a product because it is genuinely suitable or because a manufacturer has paid for preferential treatment. Retailers and technology companies will also have to establish rules governing access to consumer accounts and purchasing information.
The competition between Amazon, Walmart and artificial intelligence companies is ultimately about control over the purchasing process. The company that manages product discovery, recommendations and transaction decisions could capture a growing share of the economic value generated by online shopping.
Retailers have spent years building platforms that combine shopping, advertising and customer data. They will not willingly surrender that position to external software providers. At the same time, resisting artificial intelligence agents completely could leave them disconnected from changing consumer behaviour.
The immediate challenge is therefore not to abandon existing advertising models but to adapt them. Retailers will need to determine how external agents can access their products, how transactions are attributed and how brands can continue reaching consumers when traditional browsing becomes less important.
Artificial intelligence shopping agents may not eliminate retail advertising, but they could fundamentally change who controls it. For Amazon and Walmart, the biggest commercial risk is not necessarily losing customers. It is losing the ability to decide how those customers discover products, which brands receive visibility and where the advertising money is ultimately spent.
(Adapted from ChannelNewsAsia.com)
