Regulators look to synthetic data to tackle financial fraud

Written by Ruaraidh Gilmour and Sam Trendall on 7 October 2022 in News

Watchdogs experiment with use of artificial information to tackle fraud in which people are tricked into voluntarily sending money

Credit: Gerd Altmann/Pixabay 

Market watchdogs the Financial Conduct Authority and the Payment Systems Regulator are looking to the possible use of synthetic data to help tackle an “epidemic” of online fraud in which consumers are conned into parting with their money.

During a three-day TechSprint – often referred to as a hackathon – financial services providers, innovators, academics, regulators, and technologists were brought together to find ways to tackle the biggest problem in digital fraud.   

During the exercise the Smart Data Foundry, a spin-out from the University of Edinburgh’s £600m Data Driven Innovation programme, used its Smart Agent Simulation software to create an agent-based modelling approach to generate a synthetic data set – a process via which data that is characteristic or representative of real information is created entirely artificially.

The Smart Data Foundry said that this meant there was no commercial or privacy risk, as no real data was cloned.  

Going forward, the hope is that technique can be used to help address the growth of online bank transfer fraud known as Authorised Push Payment (APP) fraud. This refers to scams in which a person or organisation is tricked into voluntarily sending money to fraudsters.

Related content

During the Techsprint event, participants used synthetic data to simulate events such as mobile phone conversations and emails, as well as their connections with financial fraud.     

APP is reportedly now the UK’s foremost finance fraud type, accounting for 44% of all financial fraud and amounting to £583m in lost money.

Jessica Rusu, chief data, information and intelligence officer at the FCA, said: "TechSprints bring together participants from across, and beyond, the financial services sector, to develop technology-based ideas or proof of concepts that address and help provide solutions to industry challenges. These events help bring important systemic issues to the forefront, inform our thinking as a regulator, and provide a forum for collaboration and innovation. Using synthetic data helps bring to life the challenges we're working to fix and lets the TechSprint participants quickly test ideas and show the art of the possible."  

Bryn Coulthard, chief platform officer of Smart Data Foundry, said: "Smart Data Foundry has united with fintech innovators to help tackle the UK’s APP epidemic. Our high utility synthetic data allows us to understand complex scenarios without exposing customer data as we work to solve our top financial fraud problem.  Our work in partnership with the FCA and PSR helps to shine a light on the issues affecting consumers and how they are increasingly struggling to make ends meet, not just because of rising costs but also because of the ever-present risk of becoming a victim of a scam or fraud." 


About the author

Ruaraidh Gilmour is a reporter at PublicTechnology sister publication Holyrood, where a version of this story first appeared. He tweets as @Ruaraidh0.


Share this page




Please login to post a comment or register for a free account.

Related Articles

Number of services using GOV.UK Notify has quadrupled since 2020
22 March 2023

GDS reveals that uptake of messaging platform shot up during pandemic

What apps are on government’s approved list?
20 March 2023

Only centrally approved third-party applications will be allowed on Whitehall devices – but government remains tight-lipped on what might make the cut or how

MPs call for transformation plan of ‘antiquated’ DVLA
20 March 2023

Committee flags up need for IT infrastructure to be more resilient and better support remote working

Research firm brought in to improve assessment of major tech projects
17 March 2023

Government’s Evaluation Task Force has signed a deal with Ipsos to address ‘significant lack of good quality evaluation’