ChatGPT Is Now Recommending Scam Websites That Steal Your Credit Card Info
Imagine asking ChatGPT for help finding a pair of shoes, only to be directed to a fake storefront that steals your payment details. This is no longer a hypothetical scenario. According to a recent report by The Guardian, OpenAI’s chatbot is actively promoting fraudulent retail sites built to harvest credit card information. The discovery, made by scam-checking service Ask Silver, highlights a dangerous new tactic: using AI to amplify online fraud. ChatGPT scam websites now pose a real threat to unsuspecting shoppers.
How Scammers Exploit ChatGPT for Fraud
Scammers have found a clever way to manipulate AI. Instead of phishing emails or fake ads, they are poisoning the data that powers ChatGPT. Ask Silver’s Anna Jones told The Guardian that fraudsters create content designed to rank highly in the chatbot’s responses. This technique, sometimes called data poisoning, tricks the AI into recommending malicious URLs.
These fake sites mimic legitimate storefronts with official-looking web addresses. They target brands that have recently shut down or been acquired, leaving a gap in consumer demand. For example, Russell & Bromley, a British footwear retailer that went into administration in January 2026, had no official site after being absorbed by Next. Scammers built a convincing clone and optimized it to appear in ChatGPT’s search results. Shoppers asking for Russell & Bromley products by name were pointed straight to the fraudulent page.
Defunct Brands: A Prime Target for AI Scams
Fraudsters are deliberately focusing on brands that no longer have an active web presence. When a company closes or is acquired, customers often search for remaining stock or deals. This creates a perfect opportunity for scammers to step in. ChatGPT scam websites exploit this gap by offering products that seem too good to be true—often at steep discounts like 80% off.
OpenAI has since updated its results for shopping queries related to Russell & Bromley. The chatbot now displays a warning: “Several websites currently advertising Russell & Bromley products at 80% appear suspicious and may not be official retailers. Recent reports have highlighted fake Russell & Bromley stores appearing in AI-powered search results.” However, this fix is reactive, not proactive. Other defunct brands could still be vulnerable.
The Growing Problem of AI Shopping Scams
This issue extends beyond one brand. Research published last year found that ChatGPT already struggles to give consistent, reliable product recommendations. Misdirecting users to scam sites represents a far more serious failure. As AI becomes a more active part of the buying process, the risk escalates.
Louise Baxter of the UK’s National Trading Standards warned The Guardian that fraudsters adapt quickly to new technology. With agentic AI shopping still in its early stages, the window to address this vulnerability is narrow. Consumers who rely on ChatGPT for product research could unknowingly hand over their credit card details to criminals.
How to Protect Yourself from ChatGPT-Powered Scams
First, always verify the URL of any recommended site. Look for subtle misspellings or unusual domain extensions. Second, check if the brand has an official website by searching independently—not through the chatbot. Third, use a credit card with fraud protection when shopping online. Finally, report any suspicious sites to authorities like the Federal Trade Commission.
For more tips on staying safe online, read our guide on how to avoid online shopping scams. You might also find our article on best AI tools for safe shopping helpful.
What OpenAI and Regulators Are Doing
OpenAI has acknowledged the issue and is updating its safety systems. The company now includes warnings for certain queries, but this is a patch, not a permanent fix. Regulators like the UK’s National Trading Standards are investigating, but enforcement remains challenging. As AI evolves, so do the tactics of fraudsters.
In the meantime, shoppers must stay vigilant. ChatGPT scam websites are a reminder that AI is not infallible. Always double-check before clicking a link or entering payment information. The convenience of AI comes with responsibilities—both for developers and users.
