ChatGPT Shopping Scams Exploit Fake Retailer Sites

AI-powered search tools like ChatGPT are increasingly being used by consumers seeking shopping recommendations. However, this convenience is being exploited by scammers who create convincing fake websites, leading unsuspecting buyers to believe they are purchasing from legitimate retailers. These fraudulent sites are then appearing in AI-generated search results, effectively “poisoning” the information users receive.

Scam Watch, a scam-checking service, has identified instances where cloned websites mimicking popular retailers such as Russell & Bromley and Dunelm have appeared in ChatGPT’s search results. Anna Jones of Scam Watch suggests that the large language model (LLM) powering ChatGPT may have been compromised by malicious content, a phenomenon known as “poisoning.” This occurs when fraudsters insert fake content, such as cloned webpages, into the data that AI models learn from.

The vulnerability is exacerbated by factors like the administration of Russell & Bromley in early 2026 and its subsequent absorption by Next. This means that while official Russell & Bromley websites no longer exist, consumers may still search for them, creating an opportunity for scammers to capitalize on these searches with fake sites.

Fake Websites Mimic Real Retailers

The fake websites identified by Scam Watch are designed to look highly credible, often featuring substantial discounts of up to 80% to lure customers. The URLs for these fraudulent sites can be deceptive, incorporating terms like “official” or “deals” alongside variations of the retailer’s name, such as therussellbromleyofficial, russellandbromleylondon, russellbromleyonlineuk, and russell-and-bromley. Legitimate UK retail sites typically use .co.uk or .com domain extensions.

When consumers click through from ChatGPT’s recommendations to these fake sites, they are led to believe they are browsing authentic online stores. The reported scams involve users purchasing items that are never delivered, with their money and sensitive banking details being stolen. A significant red flag for these fraudulent sites is that they often exclusively accept payment via bank transfer, a method that offers little recourse for consumers in case of fraud.

Industry Response and Consumer Advice

Louise Baxter, head of the scams team at National Trading Standards, has warned consumers against assuming the legitimacy of a website simply because it is recommended by an AI tool. She stated, “Consumers are increasingly turning to AI tools for advice and recommendations, but criminals are adapting just as quickly. The fact that scam websites can appear in AI-generated results is worrying, and is a stark reminder that fraudsters will exploit any new technology that helps them reach potential victims.”

A spokesperson for Dunelm emphasized the importance of customers engaging only with their official website, www.dunelm.com, or the official Dunelm app. The company stated that it actively works to ensure the swift removal of any fraudulent sites when they become aware of them.

Next, which acquired Russell & Bromley, has acknowledged the situation and is collaborating to have the fake sites shut down. A spokesperson for ChatGPT confirmed that fraudulent websites have been removed from its search index and provided a form for users to report sites that violate their policies.

Consumers who believe they have fallen victim to such scams and have handed over financial details are advised to report the incident to their bank and to Report Fraud.

Datos clave
| Aspect | Details |
|—|—|
| Scam Type | AI-generated shopping recommendations leading to fake websites |
| Exploited AI Tool | ChatGPT |
| Affected Retailers | Russell & Bromley (cloned sites), Dunelm (cloned sites) |
| Red Flags | Deceptive URLs, excessive discounts, payment via bank transfer only |
| Advice | Verify website addresses, avoid trusting AI recommendations blindly, use official retailer sites directly |

This development is significant for ReviewArticle readers as it highlights a new and evolving threat vector in online commerce. As AI tools become more integrated into daily life, the potential for malicious actors to exploit these technologies for fraudulent purposes increases. Consumers need to be aware that AI-generated content, including shopping recommendations, is not infallible and can be manipulated. Vigilance and critical evaluation of online sources are paramount to protecting oneself from financial loss and data theft in the digital marketplace.

Fuente: https://www.theguardian.com/money/2026/jun/07/ai-chatgpt-shopping-scams-fake-websites

Datos clave