Privacy-enhancing technologies (PETs) focus of global prize announced at Summit for Democracy 2023
Privitar, a leader in data security and privacy solutions, has been announced as joint-winner of the UK-US PETs Prize Challenge, a transatlantic innovation prize challenge created by the United Kingdom and United States governments to advance the use of privacy-enhancing technologies (PETs) in tackling some of our most pressing societal challenges.
As leader of “Project STARLIT”, a collaboration with University College London and Cardiff University, the award recognises Privitar’s contribution to enabling responsible data use through the development of technologies that preserve the fundamental human right to privacy.
“Digital transformation has resulted in a massive increase in the demand to share and collaborate with data” said Jason du Preez, CEO, Privitar. “A strong data culture is critical to the success of organisations today. PETs enable data to be used without leaking sensitive information.”
In a world where AI is increasingly prominent, and the resulting appetite for high quality data insatiable, designing trust into systems is an imperative. The PETs Prize Challenge centred around privacy-preserving federated solutions that enable learning algorithms to be trained on sensitive data without the need to reveal, share, or combine raw data.
Project STARLIT: addressing practical data privacy concerns in real world scenarios
Project STARLIT builds on Privitar’s data security and privacy platform (DSP), which enables effective and responsible data use through the use of PETs. Dr Suzanne Weller, Head of Research at Privitar and lead researcher for the project explains: “PETs enable new opportunities to unlock the value of sensitive data through data sharing and collaborative analysis without compromising data privacy. The prize challenge use cases on financial crime prevention and pandemic forecasting are prime examples.”
The winning solution from Project STARLIT enables multiple financial services organisations to collaboratively train a fraud detection model without directly sharing or centralising their data. A higher accuracy can be achieved through this collaborative analysis compared with training on data from a single organisation.
PETs have a huge potential to improve trustworthy data innovation, particularly as part of a holistic data governance strategy. “This needs to be done holistically and coherently with other appropriate security measures, combining fine-grained access controls, data protection capabilities, and advanced privacy-enhancing technologies, like those showcased in this competition,” adds du Preez. “We’re delighted to win as it demonstrates our commitment to research and innovation with important use cases that are clearly relevant on the global stage.”
The award was announced at The Summit for Democracy convened by President Joe Biden, an international event that aims to strengthen democracy and defend against authoritarianism; address and fight corruption; and promote respect for human rights.
Entries to the competition went through multiple phases and were evaluated against their privacy, accuracy, efficiency, explainability, and innovation. To judge the privacy of the solutions, ‘Red Teams’ were recruited to carry out intentional attacks on the system in an attempt to reveal sensitive information about individuals.
The solution was additionally awarded a Special Recognition prize for its innovative approach to the modelling task and system architecture.
All winners of the PETs Prize challenges will be showcasing their solutions to industry innovators, government officials and regulators, and investors on 22nd May at an event organised by UK Department for Science, Innovation and Technology Centre for Data Ethics and Innovation (CDEI).
Visit AITechPark for cutting-edge Tech Trends around AI, ML, Cybersecurity, along with AITech News, and timely updates from industry professionals!
