BGL’s anti-fraud model wins major award

A new counter-fraud machine-learning model developed by a Peterborough-based finance firm has received a national award for fighting fraud.

BGL - which owns price comparison site Compare the Market - launched the model earlier this year in response to rising levels of so-called ‘ghost broking’, which sees cyber criminals misuse genuine but stolen name and address details to buy policies which they then sell on to unsuspecting third parties.

The model received the award for ‘Combating Fraud: Broker Innovation of the Year’ during a ceremony held at The Brewery in London on October 4th.

It uses machine learning to block attempted fraud at the point of quotation, while validating the 99 per cent of customer enquiries which are genuine.

At its heart is an algorithm that processes over 100 fields of customer and behaviour data into 750 model features which help isolate fraudsters based on their suspicious characteristics, picking holes in their imperfect attempts to impersonate their victim.

As a result, the volume of fraudulent policies taken out has reduced by more than 60 per cent. Elsewhere, the number of post-sale checks and balances needed to ensure that customers are genuine has dropped too.

A spokesman at BGL said: “It’s important to all of us at BGL that we keep one step ahead of fraudsters to protect the millions of genuine customers who rely on us.

“Our passionate and determined fraud and data teams play a vital role in preventing fraudulent activity through their cutting-edge techniques. This award is a testament to their innovative and creative approach.”