UCL Computer Science


UCL Computer Science Team Wins UK-US Privacy Enhancing Technologies Prize Challenge

31 March 2023

A team of academics and students from the Information Security Research Group, along with data privacy firm Privitar and Cardiff University, has been named joint first winners of the UK-US Privacy Enhancing Technologies (PETs) Prize Challenge.

A shadowy figure using a computer with code on it

The competition aims to promote the use of PETs in the financial sector to identify and prevent financial crime while preserving customers’ personal data. PETs allow for privacy-preserving financial information sharing and analytics, enabling the identification of anomalous payments without compromising individuals' privacy. 

Professor Steven Murdoch, Dr Aydin Abadi, Mohammad Naseri and Dan Ristea, who were part of team STARLIT, created a prototype system designed to detect suspicious financial transactions while also protecting the privacy of innocent individuals.  

The system employs machine learning to analyse information about the transactions and parties involved without collecting all of the data in one place. STARLIT tested the system using synthetic data provided by SWIFT. 

The PETs Prize Challenge evaluated the solutions based on their accuracy, efficiency, level of privacy protection, and innovation. 

STARLIT were announced as joint first place winners at the Summit for Democracy, alongside the University of Cambridge. The team were also honoured with a special recognition prize for their work.  

Commenting on the achievement, Professor Steven Murdoch said: “Privacy enhancing technologies show great promise in allowing the safe sharing of data, but to fully realise these benefits more research is needed. I am pleased that this competition recognises the innovation that can come from collaborations between industry and academia.” 

The Prize Challenge was part of a joint initiative between the UK and US to transform financial crime prevention and boost pandemic response capabilities through privacy-preserving federated learning. 

Read more about the PET Prize Challenge