Featurespace has secured funding from UK and US governments to develop a prototype privacy-enhancing artificial intelligence system to help banks and payments service providers (PSPs) detect financial crime.
The cash comes from the privacy enhancing technologies (PETs) Challenge Prize, an initiative launched in July 2022 by Innovate UK and National Science Foundation in the US, and supported by bank-owned messaging network Swift.
Under the scheme, Featurespace will build a prototype AI model trained on sensitive private payments data without organisations having to reveal, share, or combine their raw data. The firm says it will apply federated deep learning incorporating techniques such as k-anonymity and local differential privacy to the project.
David Sutton, Featurespace’s director of innovation, comments: “UK and US governments want banks to work together to stop fraud and money laundering. This type of privacy-preserving collaborative AI is a hard problem that no-one has yet solved. A successful outcome of this project is to make money laundering across borders and between banks much more difficult.”
Featurespace now has until 24 January to build its AI prototype. If successful, the application will be showcased at the second Summit for Democracy in the US, which President Biden plans to convene in the first half of 2023.