New Lloyds Bank system detects fraud before it happens

The fraud team at Lloyds Bank has developed a new way to spot scams before its customers lose any money.

The sophisticated software – nicknamed ‘The Rat’ – has been built in-house by experts in fraud prevention, software engineering and data science.

The “cutting-edge” technology went live at the end of last year, but the bank only revealed today that it had already prevented more than 1,900 customers falling victim to scams, which could have totalled £4m.

The technology works behind the scenes in the Lloyds internet banking portal, and detects tell-tale signs of an impending fraud through a combination of three methods.

The first of these is biometrics, with the system using behavioural analysis to build a detailed profile of how a customer normally uses internet banking, which is very difficult for a fraudster to mimic. This could be how the customer usually moves around the screen, or the time it takes them to enter their personal details.

The Rat then combines this biometric data in real-time with software which can detect if remote access is being used to operate the customer’s computer – a common tactic amongst scammers who will trick the victim into passing over control of their computer.

The third factor is something that Lloyds describes as a “secret sign” – a single trait of scammer behaviour which comes up time and again in scam cases. The bank declined to make this trait public, so fraudsters would not be able to respond or try to trick the new system.

If The Rat spots the tell-tale signs of fraud, Lloyds take action by freezing the customer’s account to prevent any funds from being lost, and then contacts the customer.

Lloyds Bank retail fraud director Paul Davis said: “We are working 24/7 behind the scenes every day to help keep our customers’ money safe, by innovating and analysing data in real-time to stop fraud happening in the first place.

“We’ve used the latest technology – biometrics, big data and artificial intelligence – long before these were even buzzwords, building our fraud-fighting armoury to give us an edge over the crooks.

“One of our latest advances has been using these three detection tactics together,” he continued. “Individually, they are all good at spotting fraudsters, but when combined we can spot more fraud more quickly, and disrupt fewer genuine transactions.”

Overall losses due to authorised push payment scams totalled £354.3m across the UK last year, according to UK Finance, with one in 10 Britons reporting they have been the victim of a financial scam at some point in their lives.

    Share Story:

Recent Stories

Deep Neural Networks for FX Prediction
Adam Cadle speaks to Richard Turner Head of Research and Mike Emambakhsh, Ph.D. Senior Research Scientist at Mesirow Currency Management about their work with Machine Learning, specifically Deep neural networks for FX prediction.


Subscribe to our newsletter to receive breaking news and other industry announcements by email.

  Please tick here to confirm you are happy to receive third party promotions from carefully selected partners.