Study finds AI chatbots sometimes point shoppers to fake or risky retail websites

A new study has found that popular AI chatbots sometimes send shoppers to fake or otherwise risky retail websites. According to The Verge, which received the results exclusively, the research was carried out by IP protection firm Podqi and covered ChatGPT, Gemini and Perplexity.

Podqi asked the chatbots shopping-related questions about 151 small and large brands, including Lululemon, Birkenstock, Stüssy, Madhappy and Le Creuset. Madhappy is a client of Podqi. Sample prompts included where to buy a brand's products, where to find it online, and whether it ships internationally. The tests ran in two periods in September, and Podqi wrote the prompts itself rather than using real user conversations.

Across 6,795 answers, the firm counted a site as risky if it was a fake page imitating a genuine retailer or selling knockoffs, an old brand domain re-registered by someone else, or an inactive site that another party could buy and reuse. All three chatbots surfaced such sites at least some of the time.

Perplexity fared worst, showing a risky site in 15 percent of its answers. ChatGPT came in at 8 percent, and Google's Gemini did best at 4 percent. Podqi counted mentions of any kind, including recommending a site, showing it without comment, and warning the user about it. ChatGPT was the most likely to include a warning, doing so 24 percent of the time.

Smaller brands looked more vulnerable. Podqi found that questions about larger companies produced risky sites about half as often. Questions about regional brand websites also triggered more risky results than other prompt types, and the chatbots sometimes listed fake regional sites for brands that have none. In one case, Perplexity named the wrong page as Stüssy's official site and described it as the best starting point for guaranteed authentic items.

The report notes that the findings could matter ahead of the holiday shopping season, when more people are expected to use AI to hunt for in-stock items and good prices, possibly from retailers they do not know. Podqi CEO and cofounder Trevor West said consumers should verify the source of a result before entering personal or credit card information.

The companies responded by email. Google spokesperson Jessica Levis Boord said Gemini displays links to sources so people can check information, and that the company is open about the chance of mistakes. OpenAI spokesperson Taya Christianson said the company is improving safeguards to make deceptive sites less likely to appear in shopping results. Perplexity spokesperson Jesse Dwyer said the company never claims to be fully accurate but focuses relentlessly on accuracy.