Starling Bank Introduces New AI Tool to Combat Purchase Scams
Starling Bank has launched a new artificial intelligence tool designed to assist its customers—whether they have personal, joint, or business accounts—in identifying potential purchase scams.
The innovative tool, named Scam Intelligence, allows customers to upload images related to their purchases, including photos of items, listings, and conversations with sellers on online marketplaces. By using this feature, customers can receive advice on whether the transaction might be a scam. Scam Intelligence can also evaluate screenshots of texts and messages from individuals asking Starling customers to send money.
For instance, if someone is considering buying a bicycle from Facebook Marketplace, Scam Intelligence can alert them if the price seems suspiciously low, if the image in the ad appears fake, or if the bank account details do not match those of the seller.
Similarly, if a user wants to purchase a guitar from eBay, the tool may flag instances where the seller avoids using eBay’s secure payment options. Equipped with this information, customers can make more informed decisions about whether to proceed with their purchase.
Scam Intelligence is powered by Google’s Gemini models and operates on the Google Cloud platform. This technology helps the tool understand the context of the images and texts that customers upload, allowing Starling’s systems to provide an accurate risk assessment.
Catherine Britton, head of fraud risk at Starling Bank, stated, “We’re on a mission to help our customers be ‘good with money’, from spending to saving and protecting their funds. While we collaborate with social media platforms and mobile networks to tackle fraud at its origin, Scam Intelligence is an additional tool our customers can use to safeguard themselves.”
Fraud has become a significant issue in the UK, costing the country £1.17 billion in 2024. Of this amount, authorised push payment (APP) fraud—where customers unknowingly approve payments to scammers—accounted for £450 million.
