One fake web page can be enough to trick AI shopping recommendations

# AI Shopping Assistants Can Be Tricked Into Promoting Fake Products When you ask an AI chatbot for shopping recommendations, it searches the web and uses those results to give you suggestions—but researchers discovered that just one fake webpage is enough to fool it into recommending made-up brands as if they were real. The problem isn't hypothetical; it's already happening, with fake brands managing to reach the top of AI recommendation lists within hours. This matters because AI now drives about 2% of all traffic to major retailers like Target and Walmart, meaning these recommendation tricks could actually influence where you shop.
AI recommender systems are becoming an increasingly important part of how we shop. About 2% of all referrals to major shopping websites like Target and Walmart come from large language models, according to data.ai. But LLMs can be easily swayed, according to a new study by Minghao Luo and Liang C
More from Best AI Tools
Get new guides every week
Real AI income strategies, tool reviews, and plain-English news — free in your inbox.



