Scammers Plant 80% Off Fake Retail Sites in ChatGPT Results
Updated
Updated · The Guardian · Jun 7
Scammers Plant 80% Off Fake Retail Sites in ChatGPT Results
3 articles · Updated · The Guardian · Jun 7
Summary
Ask Silver found cloned Russell & Bromley and Dunelm websites appearing as sources in ChatGPT shopping answers, exposing users to payment theft and stolen bank details.
Two fraudulent Russell & Bromley sites surfaced when researchers asked for popular bags, suggesting the model had been “poisoned” with malicious cloned webpages.
Russell & Bromley’s January 2026 administration created an opening: the brand now sits within Next’s website, but shoppers may still search for a standalone store and land on fakes.
ChatGPT said it removed the fraudulent sites from its search index, while Next said it was working to shut them down and Trading Standards warned AI recommendations are not proof a site is genuine.
Ask Silver urged shoppers to go directly to retailers’ sites, avoid domains with extra words like “official,” and treat bank-transfer-only payments and steep discounts as red flags.
Are AI models being deliberately 'poisoned' by scammers, or just learning from a fraudulent web?
Beyond playing whack-a-mole with fake sites, how can we immunize AI from malicious data manipulation?
As AI-driven scams surge, who is legally liable for the losses: the AI developer or the user?
The 2026 AI Scam Epidemic: 3,000 Daily Complaints, Business Risks, and How to Shop Safely Online
Overview
In 2026, the digital world faces a sharp rise in online scams, driven by the advanced capabilities of artificial intelligence. AI makes scams more convincing and harder to detect, leading to serious financial and reputational risks for both individuals and businesses. The Internet Crime Complaint Center now receives nearly 3,000 complaints daily, showing how widespread these threats have become. Identity fraud has become a major concern, as AI tools help scammers create fake documents and realistic schemes. This new wave of AI-powered scams is changing how people and organizations must protect themselves online.