Data Trusts: What Are They There To Do, and What Should They Look Like?

Data Trusts: What Are They There To Do, and What Should They Look Like?

Abstract: In Wendy Hall's 2017 AI review, the first recommendation was to create data trusts to encourage data access and data sharing. A number of initiatives are exploring these ideas in concrete contexts, including Elena Simperl's Data Pitch project and the Open Data Institute's work on data innovation. This talk will consider the question in more abstraction, to consider data trusts' function, and how this will determine their form. I will argue that a data trust should work within existing law to provide ethical, architectural and governance support for trustworthy data processing, and I will unpack this in terms of how it might create trust, what architectures might help implement data trusts, and how they relate to the existing law of trusts. Biodata: Kieron O'Hara is an associate professor in the WAIS group of ECS. His research interests are in trust, privacy and the nature of digital modernity. He is a lead in the UKAN network of anonymisation experts. His latest book, The Theory and Practice of Social Machines (with Nigel Shadbolt, Dave De Roure and Wendy Hall) will be published by Springer in March.