When AI shares incorrect information, 58 percent of shoppers blame the brand

It’s no big secret that artificial intelligence is changing the face of retail. Applicable at nearly every stage of the broader product lifecycle, more and more consumers across the globe are using large language models (LLMs) from ChatGPT to Gemini for online shopping. Turning to AI to discover, compare, and select which products from which brands are worth the purchase, LLMs are increasingly becoming an influential part of the buying experience, one that fashion brands and retailers would be wise to invest in.

To better understand how AI is impacting online commerce, cloud-based commerce platform Rithum and Studio’s Retail Dive carried out a survey among 1,046 US and UK consumers, publishing the findings in the report ‘The New Discovery Engine’. Here we outline some of the main takeaways that outline the potential LLMs offer in terms of reaching and connecting with the right audience. As long as your brand’s content, pricing, and inventory data are correct, the potential to attract more customers is high. But if the data falters, then be warned, as AI will offer its own story about your brand, whether it’s true or not, and feed this to shoppers.

Consumers (including your target audience) are already using AI to shop

Adoption of LLMs for shopping among consumers in the US and UK is common, but it does vary by age and income. For example, the report found that more than 8 in 10 shoppers under the age of 44 have used an LLM as part of their shopping journey in the last three months, with over half stating they trust AI tools as much as they trust brands and retailers' own websites.

Unsurprisingly, younger demographics were found to be the largest group using and relying the most on AI. The report found that 80 percent of 18 to 43-year-olds had used AI for shopping during the last three months, compared to 51 percent of the 60-and-older age group. Many of the younger consumers added they would miss out on the benefits offered by using AI for shopping, with more than one in four of 28 to 43-year-olds stating they would feel “a lot” of loss if they were not able to access AI tools for shopping, compared to 9 percent of those ages 60 and older. 

Most shoppers surveyed stated they use AI mainly for research purposes, to explore product details and compare options, with over 90 percent relying on it for these tasks. More than half also said they use AI to decide where to buy. Rather than letting AI make the final purchasing decision, AI helps shoppers narrow choices by surfacing top options that guide which products, retailers, and brands they ultimately consider. 37 percent of those surveyed also noted they used AI when shopping for apparel, shoes, and jewelry, slightly less than electronic purchases (45 percent). 

Not all consumers use AI the same, with higher-income households relying more on AI

Household income is another aspect fuelling AI adoption. The survey found that higher-income households, aka those with more purchasing power, use AI for shopping more, with usage reaching 84 percent among households earning between 100,000 and 150,000 US dollars annually. Households with a lower income, those earning under 30,000 US dollars, saw AI usage drop to 56 percent. 

Income differences were also found to impact how shoppers use AI, with lower-income households using AI as a price tool to secure the best deals. In fact, 43 percent of households earning less than 30,000 US dollars mainly use AI to “find the best price” when shopping. Higher-income households, on the other hand, see and use AI to help reduce time spent shopping and avoid having to browse multiple sites to find the product they need. They’re also twice as likely to trust AI without visiting another site.

AI is helping small brands beat out household names

AI tools are helping shoppers feel more confident when shopping, but they are also making consumers less brand loyal. 43 percent of respondents compared more product options with AI, with 36 percent using it to make faster buying decisions, and 34 percent feeling more confident about their purchases. 

Connecting shoppers with brands they may not know, 19 percent or close to 1 in five shoppers said they had purchased from a brand they’d never previously heard of, because AI recommended it. 13 percent also said that they’re more likely to switch retailers, brands or products based on AI suggestions and 32 percent spend less time browsing other websites when using AI. Unsurprisingly, the shoppers most likely to act on those AI recommendations without additional research or a second option were power users and high-income shoppers.

When AI gets it wrong, your brand takes the blame

While LLMs are a powerful aid in product discovery, the accuracy of product data, from prices to availability, materials, sizing, and other specifications, is just as important. 53 percent of survey respondents were found to trust AI tools as much as brand sites. 

What AI shares about your brand and product is incredibly important, and when AI provides bad or incorrect product information, consumers don’t just blame the LLMs. In fact, 58 percent of consumers stated that their overall trust in a brand or product decreased when an LLM serves up wrong information, with 16 percent abandoning the purchase altogether. 

Price accuracy matters most in AI recommendations

Price was found to be the most critical factor in AI-driven recommendations, with 67 percent of shoppers prioritising its accuracy, followed by product reviews (35 percent) and availability (34 percent). However, 44 percent of consumers surveyed also noted that AI tools still need to improve the accuracy of core product details, particularly pricing and stock levels.

At the same time, shoppers also rarely verify AI-generated information on brand websites, with only 5 percent of those surveyed doing so, suggesting decisions are largely formed before reaching owned brand channels. Instead, they rely on search engines (28 percent), friends and family (17 percent), or prior experience (17 percent), while 64 percent of 18–27-year-olds purchase based on AI recommendations without further validation.

“AI is making consumer trust a product data problem. As agentic commerce becomes the baseline for how shoppers research and evaluate products, brands need to think more strategically about how their product information appears across their entire commerce ecosystem,” said Sam Griffin, VP, Strategy and Engagement at Rithum, in a statement. “Accurate, consistent product data will play a major role in how brands are discovered by AI and trusted by consumers in agentic shopping experiences.”