New anti-fraud tool as Scottish data experts launch groundbreaking synthetic dataset

University of Edinburgh spin-out Smart Data Foundry (SDF) is to partner with key financial bodies to use groundbreaking synthetic dataset against fraud.  

Earlier this year, SDF unveiled its engine, Aizle, which looks to provide UK innovators with synthetic data, aiming for a safer, easier to use and more accurate output. 

Working with the Financial Conduct Authority (FCA) and the City of London Corporation, it has used Aizle to create a new dataset to tackle Authorised Push Payment (APP) fraud. 

APP fraud scams happen when someone is tricked into sending money to fraudsters posing as regular people. Last year, almost £500m was lost to APP scams, accounting for 40 per cent of the overall financial fraud.  

The material can be accessed via the FCA’s Permanent Digital Sandbox. 

Formally introduced in August, the sandbox is a testing environment which allows firms to experiment with their products through 'proof of concepts'. The tool also encourages data providers to list their information for a better understanding and traffic on its use.  

This detailed insight into the end-to-end lifecycle of APP fraud, which is often committed against the financially vulnerable, has been long-awaited so researchers can test ideas and models.

This partnership continues the effort from last year’s FCA and Payment Systems Regulator (PSR) TechSprint event, which focused on the issue and saw SDF's first-ever use of the APP Fraud synthetic dataset. 

Bryn Coulthard, chief product and technology officer at SDF said “We focus on creating high utility synthetic data to enable innovation within the financial services industry. We are delighted to continue partnering with the FCA with our APP fraud datasets to help play a part in tackling this growing problem and to help ignite and accelerate innovation in this space.”