Artificial intelligence tools like ChatGPT have revolutionized how we shop online, offering personalized recommendations and instant product research. However, a new wave of sophisticated scams is exploiting these AI assistants, leading unsuspecting consumers directly to fake websites that steal their money and personal data. Recent investigations by scam-checking service Ask Silver have uncovered a troubling trend: cloned retail sites are appearing in ChatGPT's search results, tricking users into believing they are shopping with trusted brands.
The scam works by a process known as AI poisoning, where malicious content is inserted into the information a large language model (LLM) learns from. Fraudsters create convincingly realistic copies of legitimate retail websites, and when users ask ChatGPT for product recommendations, the AI may unwittingly cite these fraudulent sites as sources. This creates a dangerous illusion of authenticity, as consumers trust the AI's judgment over their own scrutiny.
How the Russell & Bromley Scam Unfolded
One prominent example involves the luxury footwear and accessories brand Russell & Bromley. After the company went into administration in January 2026 and was absorbed by Next, its official standalone website disappeared. Scammers quickly exploited this vacuum. Ask Silver's research revealed that when users asked ChatGPT about popular Russell & Bromley bags, the AI listed several product options with prices and links. Among the cited sources were at least two fraudulent websites, including addresses like therussellbromleyofficial and russellandbromleylondon.
These fake sites were meticulously crafted to look identical to the original brand's online store, often offering massive discounts of up to 80% to lure in bargain hunters. When a customer makes a purchase, the scammer simply takes the money and never ships the product, while also harvesting sensitive bank details for future fraud. The legitimate Russell & Bromley store now only exists within the Next website, making the cloned sites the only visible option for many shoppers.
Why AI Is Vulnerable to This Fraud
Large language models like ChatGPT generate responses based on the vast corpus of text they have been trained on, which includes the entire public internet. If scammers create and index hundreds of fake webpages, the AI can inadvertently treat them as credible sources. Anna Jones of Ask Silver explains that the LLM may have been poisoned by these malicious sites, especially when a brand's official online presence disappears, leaving a gap that fraudsters rush to fill.
Louise Baxter, head of the scams team at National Trading Standards, warns that consumers should never assume a website is genuine just because it is recommended by an AI tool. She states, "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." This highlights a critical vulnerability in the current AI ecosystem: it has no built-in mechanism to verify the legitimacy of commercial links.
How to Spot a Cloned Website
Protecting yourself from these scams requires a sharp eye and a healthy dose of skepticism. Below is a quick reference table of red flags to watch for when shopping via AI recommendations:
| Red Flag | What to Look For |
|---|---|
| Suspicious URL | Extra words like "official", "london", or "uk" in the domain; missing or unusual TLDs (e.g., .shop instead of .co.uk) |
| Unrealistic Discounts | Offers of 70–80% off popular items, especially on luxury brands |
| Poor Design or Grammar | Mismatched fonts, broken links, or awkward product descriptions |
| No Secure Checkout | Payment pages without HTTPS (lock icon) or using unverified payment processors |
| Brand Website Moved | If the brand was recently acquired or went into administration, verify where they now sell |
Always cross-check the URL directly. Legitimate UK retailers typically use .co.uk or .com domains without extra descriptive words. If a deal seems too good to be true, it almost certainly is. Before entering payment details, visit the brand's official social media pages or contact customer service to confirm their current web address.
FAQ: AI Shopping Scams
What is AI poisoning in the context of shopping scams?
AI poisoning is when fraudsters deliberately insert malicious or misleading content into the data that an AI model learns from. In the case of ChatGPT, scammers create fake websites that are indexed by search engines, and when the AI is asked for recommendations, it may cite these fraudulent sites as trustworthy sources, leading users to scam pages.
How can I safely use ChatGPT for shopping recommendations?
Always treat AI recommendations as a starting point, not a final verdict. Manually verify any website by searching for the brand name separately, checking for official social media accounts, and looking for secure checkout indicators (HTTPS). Never click directly on links provided by an AI without first inspecting the full URL for suspicious elements.
What should I do if I think I've fallen for a fake website scam?
Immediately contact your bank or credit card provider to report the transaction and freeze your cards. Change passwords for any accounts you used during the purchase. Report the scam to your country's consumer protection agency, such as the National Trading Standards in the UK or the Federal Trade Commission (FTC) in the US. You should also monitor your bank statements for any unauthorized charges.
