The Hidden Dangers of AI Shopping Assistants: A Warning on Fraudulent Links

The rapid integration of AI chatbots into the online shopping experience has introduced a new layer of risk for consumers. A recent study conducted by the intellectual property protection firm Podqi has revealed that popular AI tools, including ChatGPT, Gemini, and Perplexity, are frequently directing users toward fraudulent or highly suspicious retail websites. This discovery raises significant concerns about the safety of using generative AI as a primary tool for product research and price comparison.

In the study, researchers prompted these AI models with various shopping-related queries. Instead of providing verified links to legitimate retailers, the chatbots often hallucinated or scraped data from unreliable sources, leading users to impostor sites designed to steal personal information or financial data. These sites often mimic the branding and aesthetic of reputable stores, making it difficult for the average shopper to distinguish between a genuine storefront and a sophisticated scam.

This issue highlights a fundamental flaw in how current large language models process and retrieve information. Because these models are trained on vast amounts of internet data, they can inadvertently prioritize search results that are optimized for visibility rather than legitimacy. As AI assistants become more integrated into browsers and search engines, the potential for widespread consumer harm grows. Experts are now calling for better verification protocols and stricter data filtering for AI models that provide commercial recommendations. Until these safeguards are implemented, consumers are advised to exercise extreme caution when clicking on links provided by chatbots and to verify the authenticity of any retail website before entering payment information.

Sources: The Verge