A brand-protection study says roughly 9% of links returned by three AI shopping assistants led to risky sites, including fake shops and inactive domains. The findings are a useful prompt for Black Friday shoppers: treat chatbot recommendations as leads to check, not shops to trust.
Watch Desk analysis
What happened
BigGo Finance reports that Podqi asked ChatGPT, Gemini and Perplexity shopping questions about 151 brands, then analysed 6,795 responses. About 9% pointed to risky sites. Perplexity had the highest reported rate at 15%, followed by ChatGPT at 8% and Gemini at 4%.
Of 718 flagged domains, the report says 351 were fake shops and the rest were inactive. More than half of the risky links were presented as buying recommendations, while only 9% came with a warning. Questions about smaller brands and retailers in other countries produced the highest error rates.
Why it matters
A chatbot can produce a convincing shortlist and still send a shopper somewhere they should not enter payment details. The study measured links to risky destinations, not whether anyone was defrauded, and its results should not be read as a measure of every assistant’s performance in every market.
The practical steps are refreshingly unglamorous: ask for links to a manufacturer’s official shop or a well-known retailer, check the live price and address yourself, and be wary of unfamiliar shops or payment by bank transfer only. For branded goods, check the manufacturer’s list of authorised sellers.
Our read
AI can help narrow the hunt, but it has not earned the keys to the checkout. A polished recommendation is not a trust mark, and a link that looks plausible can still lead somewhere decidedly less festive. Use assistants to find options; verify the shop and price before buying.
What to watch
- Whether chatbot makers publish clearer warnings when recommendations point to unfamiliar retailers.
- Whether the reported error rates change as AI shopping tools become more widely used.
- Whether shoppers verify links and prices before entering payment details.
Discussion spark: Should chatbot makers be responsible for warning users when shopping recommendations point to risky sites, or is it ultimately the shopper’s job to verify every link?
Sources and evidence
- AI Chatbots as Black Friday Shoppers' Aides: 1 in 11 Links Tied to Risky Sites, Study Finds – BigGo Finance (20 November 2026, 09:35 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.