New Rithum report finds 58% of shoppers lose trust in the brand when LLMs provide wrong information (March 25, 2026)

New survey of U.S. and U.K. shoppers reveals widespread use of AI tools for product research, rising expectations for accurate pricing, and the risk when AI recommendations go wrong

ATLANTA, March 25th, 2026 - Rithum today released a new research report, The New Discovery Engine: How consumers are using AI to find, trust, and choose brands, and what's at risk for those they never see, based on a consumer survey examining how shoppers are using LLMs and how AI search is influencing their behavior. The survey of 1,046 online shoppers across the U.S. and U.K. finds that while AI is becoming a common tool for product discovery, errors in AI-generated recommendations can directly impact brand trust. Nearly six in ten shoppers (58%) say they blame the retailer or brand when an AI recommendation contains incorrect product information, and 16% say they would avoid purchasing the product entirely after a bad recommendation.

Using AI for shopping is already widespread. According to the report, 70% of respondents have used an AI tool for shopping-related activities in the past three months, and 36% say they regularly use AI to discover new brands or products, with 43% comparing more options as a direct result of using AI. However, the technology is still primarily used during the research stage rather than the transaction itself, with fewer than 15% of shoppers reporting they have completed a purchase directly through an AI tool.

"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. "Accurate, consistent product data will play a major role in how brands are discovered by AI and trusted by consumers in agentic shopping experiences."

Other key findings from the report include:

AI is helping small brands beat out household names

• 1 in 5 shoppers have purchased from a brand they'd never previously heard of, because AI recommended it.

• Over 90% of AI-active shoppers use the tools to research products or compare options.

• 64% of 18 - 27 year olds buy based on an AI recommendation without verifying it anywhere else.

• 13% say they're more likely to switch retailers or products based on AI suggestions.

Price accuracy matters most in AI recommendations

• 67% of shoppers say price is the most important factor AI must get right when recommending products, while product reviews (35%) and availability (34%) rank behind price in importance.

• 44% say AI tools most need to improve accuracy for product details like price and availability.
  

Shoppers still verify AI recommendations elsewhere

• The verification window is nearly nonexistent as only 5% go directly to retailer or brand websites to verify product information - showing that by the time a shopper reaches an owned channel, they're already forming decisions.

• 28% of shoppers turn to search engines first to verify AI recommendations.

• Friends and family (17%) and prior experience with a product (17%) also are common verification sources.
  

As AI tools become a more prominent part of the shopping journey, Rithum's research highlights the growing importance of accurate, consistent product data across the commerce ecosystem. Brands that reliably maintain pricing, availability, and product details across channels will be better positioned to build trust and remain competitive as AI-driven discovery continues to evolve.

Check out Rithum's report, The New Discovery Engine: How consumers are using AI to find, trust, and choose brands, and what's at risk for those they never see, here.

Methodology

The findings are based on a survey of 1,046 online shoppers in the U.S. and U.K. conducted by Rithum and Studio's Retail Dive. Respondents were evenly distributed across age groups, including 18-27 (25%), 28-43 (26%), 44-59 (25%), and 60 or older (24%).