The Verge AI reports Podqi test found chatbots surfaced risky shopping sites

A Podqi study shared exclusively with The Verge found that shoppers using AI chatbots could be sent to imposters and other risky websites. The IP protection firm tested ChatGPT, Gemini and Perplexity with shopping questions about 151 brands across two September testing periods, producing 6,795 answers. All three engines cited or displayed risky sites, including fake retailers, re-registered old domains and dead sites. Gemini did so least often at 4 percent, ChatGPT at 8 percent and Perplexity at 15 percent.