Podqi Study Finds AI Chatbots Recommend Counterfeit Retail Sites

On November 13, 2026, The Verge reported on a new study from IP protection firm Podqi that examined how AI chatbots answer shopping-related questions. The study tested ChatGPT, Gemini, and Perplexity and found that in as many as 15 percent of conversations the chatbots displayed URLs flagged as risky, which could steer shoppers toward imposter sites or other dangerous pages. According to The Verge, Podqi's testing spanned 151 brands and was carried out in September, organized into two separate sessions. Podqi prompted the three chatbots with a set of shopping questions drawn from a selection of those brands, and the findings were shared exclusively with The Verge ahead of publication. Based on the available material, the public record stops at these headline results and the basic test setup. No further breakdown is provided, such as per-product risky URL counts for each individual chatbot or the full list of questions used in the evaluation.