7 in 10 US consumers use AI to shop, reshaping how Americans make their purchases

Artificial intelligence continues to reshape consumer behavior, with nearly seven in ten Americans using AI to shop for products and services, according to a new survey of 2,000 US adults from integrated marketing agency LDWW.

71 percent of those surveyed said they expect to rely on AI even more in the years to come, as AI becomes a mainstream shopping channel. Almost two-thirds of respondents said AI has influenced a recent purchasing decision, while 30 percent have used it to complete a purchase, in some cases without ever visiting a retailer's website.

Nevertheless, although US consumers are quickly adopting AI as a shopping resource, they are still auditing its responses. Nearly one in five respondents said AI tools are "the most trustworthy source" they use when shopping, with 43 percent saying they trusted the brand information AI tools share. However, more than half actively checked the citations AI provides, and almost 60 percent stated that knowing which sources inform AI responses was important to them, a "trust but verify" stance reminiscent of how consumers learned to read online reviews before making a purchase.

How AI continues to reshape the retail landscape for consumers and retailers

Other reports point in the same direction. For example, Adobe's March 2026 survey of more than 5,000 US consumers put AI shopping usage at 39 percent, but its behavioral data leaves little doubt about the trajectory. Adobe Analytics, which tracks over one trillion visits to US retail sites, found AI-referred traffic grew 393 percent year-over-year in the first quarter of 2026, after a 693 percent surge during the 2025 holiday season. The quality of that traffic has also changed: AI-referred visits converted 38 percent worse than other channels in March 2025, but 42 percent better a year later, with revenue per visit running 37 percent above non-AI traffic. In the span of one year, AI shoppers went from window shopping to outspending everyone else.

The commercial implications of agentic commerce, where AI acts on the shopper's behalf, are considerable. McKinsey projects that 750 billion US dollars in consumer spend will flow through AI-powered search by 2028, while its AI Discovery Survey found half of consumers already use AI-powered search today. However, a separate study by Cognizant and Oxford Economics found 75 percent of consumers remain unwilling to let AI complete high-value purchases autonomously, a reminder that although it may be booming, agentic checkout still has a trust ceiling.

For apparel retailers and fashion brands, the AI shift appears to be more structural rather than tactical. A previous survey by commerce platform Rithum and Studio ID's Retail Dive, "The New Discovery Engine," found 37 percent of shoppers had already used AI when buying clothing, footwear, or jewelry, and that AI can impact brand loyalty, with nearly one in five purchasing from a brand they had never heard of on an AI recommendation. However, that impact risks cutting the other way, as the survey found that when AI tools share incorrect product information, 58 percent of consumers say their trust in the brand, not the LLM, decreases, with 16 percent abandoning the purchase altogether.

A sharp reminder that sharing accurate content can make the difference in a sale or no sale. Adobe found roughly a third of retailers' product page content is unreadable to LLMs, meaning many retailers and brands are invisible to AI systems due to simple technical oversight.

Ultimately, how far AI actually impacts the final purchasing decision remains an open question. While LDWW's survey found that three in ten consumers complete purchases through AI, Rithum's data suggests consumers are using AI to help narrow a shortlist of products, making the final purchase decision themselves. Looking at the data as a whole, however, the rise of agentic commerce is undeniable. AI is steadily becoming a storefront, leaving it up to brands and retailers to figure out what merchandising for LLMs looks like.