Building Trust in Data Trusts - A talk with Jessica Montgomery

As we reach the end of our recent exploration into data ethics, we're trying to come up with some solutions and recommendations for innovative ways forward. As part of this, we've been looking into forms of data institutions, such as data commons, data unions, and data trusts. These concepts try to maximise the control and choice individuals have over their data by creating collective organisations of individuals who want to empower their data for themselves. Existing data regulations already provide rights, but the difficulty comes in exercising them. That's why these collectives, combining consenting individuals with professional help, can perhaps solve some problems.

In this vein, we've been reaching out to those working in positive and proactive directions in relation to data ethics. We've already interviewed Sam Gilbert, author of Good Data, whose optimistic and forward-thinking outlook is very refreshing. Our next interviewee is Jessica Montgomery, Director of the Data Trusts Initiative, based in the Department of Computer Science and Technology at Cambridge University. She told us more about data trusts, where that idea comes from, and how it can make a real difference.

To start out Jessica pointed us towards the research done by Sylvie Delacroix and Neil Lawrence, which explores the data governance landscape and identifies stewardship gaps in relation to empowering data subjects and maximising their ability to exercise their rights. While important, individual consent only goes so far in enabling such stewardship - it assumes that individuals have time, knowledge, and energy to investigate lengthy terms and conditions and consider how data might be shared and re-used, and it doesn't account for the relational aspects of data sharing. In short, new approaches to stewardship are needed if data sharing is to make a truly positive impact.

Data Trusts are inspired by trust law. In the UK, trusts have historically been used as a legal framework through which one person manages an asset for the benefit of another individual or group. Acting for the benefit of these 'beneficiaries' is the trustee's fiduciary duty. That fiduciary duty demands the trustee acts with undivided loyalty and creates a set of strong safeguards around how the asset held in trust is managed.

The idea behind data trusts is that people could combine their data (or data rights) in a trust and a trustee would take on a fiduciary duty to manage those resources in the best interests of the trust's beneficiaries. This is an appealing idea - it combines elements of collective action, with independent stewardship and strong institutional safeguards against data misuse - but it's still a relatively new concept so there are still some practical issues. To fill these knowledge gaps, Jessica says we need to move from theory to practice and trial different ways of working. She continues that it's especially important to learn from the nascent data trusts community, identifying their needs, and the operational strategies that implement core data trusts ideas.

So, from a policy perspective, what frameworks need to be in place to set up a data trust? Jessica says that the community is still exploring what the role of the government is in the data trust ecosystem. To develop thinking in this area, we can think about what different types of 'market failure' act as a barrier to data trusts becoming common practice in data stewardship. For this, we need to get more projects up and running. This way we can learn from practice and develop more future proof governance initiatives. We can also learn from how other countries are approaching the development of legal frameworks for data stewardship. Jessica highlighted current work by the Global Partnership on AI and the Aapti Institute into such building blocks of data trusts policy frameworks, which is exploring the rights, accountability structures and fiduciary duties involved.

But how do we get people involved without the fear of uncertain outcomes in this slightly uncharted territory? According to Jessica, while we do need to offer some guidance into the world of data trusts, this should be of a facilitative nature. Since this field is still relatively new we should support the community and learn from it. That might also require further efforts in communications and engagement: we know that public awareness of data use - and the implications of technologies like machine learning - is variable, and trusts themselves aren't an easily accessible legal framework. Efforts like the Data Trusts Initiative are needed to support those interested in creating data trusts that serve their communities and in growing the data trust community.

Jessica concludes that there is still much work to be done to support the growth and trust in data trusts but the field is developing at pace and common understandings of core concepts and ways of working are emerging. Like Jessica, we believe in supporting communities of people who believe in making the most of their data while keeping it safe. For this reason, we will keep researching modern solutions to data ethics problems, like data trusts. If you want to find out more or join us on our data ethics journey check out our project page or reach out to us directly via susanne.schmidt@open-innovations.org.