A study by IP-protection firm Podqi found that ChatGPT, Gemini and Perplexity can point shoppers to impostor or high-risk websites when asked where to buy brands, undermining trust in AI-assisted shopping.

1 min read

Study: AI Chatbots Recommend Fake Retail Websites to Shoppers

What happened

FAQ

What did the Podqi study find about AI chatbots and shopping?

It prompted ChatGPT, Gemini and Perplexity with shopping questions about 151 brands and found answers could direct users to impostor or risky websites instead of the brand's official site.

Does this affect shoppers in the Middle East and North Africa?

Yes. Arabic-speaking users asking a chatbot where to buy a global brand may receive an unofficial link, especially where awareness of domain-verification tools is limited.

How can I check whether a chatbot-suggested site is official?

Confirm the domain through the brand's own site or verified accounts, look for security certificates, and avoid shortened links or domains that only slightly resemble the real name.

Are the AI companies fixing this?

The study shows the issue across all three systems but does not include a published commitment from the companies to fix it, so human verification remains necessary for now.

Source: The Verge AI

AI-assisted summary of the source above. High-confidence stories publish automatically; others wait for review. Facts that are not in the source are not added.