Guest post by Professor Genevra Richardson, Co-chair of the British Academy and Royal Society project on Data management and use: Governance for the 21st Century.
We have reached a critical moment in the development of data science and data-enabled technologies. The growth in scale of data collection, and growth in power of analytics, means that new and previously unforeseen uses of data become possible and even routine. Data management and use: Governance for the 21st Century, an interdisciplinary report produced by the British Academy and the Royal Society, tackles the governance challenges posed by this rapidly evolving data environment.
Connecting debates on data
The Royal Society’s report Machine learning: the power and promise of computers that learn by example explored the potential of this technology to create new insights from data.
Our report on governance of data management and use looks across a wide range of sectors and disciplines, across which data is used in diverse ways and for a range of purposes. It was clear to us that common issues emerged across this landscape, issues that go beyond well-examined questions of privacy and security of data.
For example, using data relating to individuals and communities to provide more effective public and commercial services, while not limiting the information and choices available, may create concerns about transparency of decision making and about individual autonomy. The value that can be derived from data raises questions about the arrangements for data trade and transfer, and how we can balance the opportunities to society of using data against potential harm to individuals and communities.
Yet across the sphere of data management and use, common governance concepts such as consent and ownership of data are challenged by the rapid reuse and repurposing of data – requiring new approaches to governance.
A framework for data governance
These concerns will surface in diverse areas of practice, and while there is a need for specific forms of governance for particular applications, the interlinked and networked nature of data use requires a holistic approach to governance. There is a case for a common framework underpinning data governance to help anticipate what changes are required and ensure that all areas of data management and use are adequately connected and governed – both to protect people and organisations whose data is used, and to ensure that the benefits of better use of data can be realised.
This common framework is established by embedding a principled approach to governing data use. And the core, overarching principle is that systems of governance must promote human flourishing. The key idea behind this is that, while we all create data trails through our everyday activities that can be mined for insight and value, humans must never be seen as serving data and its uses – rather the use of data must always serve human communities. To deliver this core principle, systems of data governance must:
- protect individual and collective rights and interests
- ensure that trade-offs affected by data management and data use are made transparently, accountably and inclusively
- seek out good practices and learn from success and failure
- enhance existing democratic governance.
Embedding these principles across the governance landscape will require oversight and our report recommends the creation of a new kind of body – which we call a data stewardship body – guiding governance across all spheres of data use.
The stewardship body would be expected to conduct inclusive dialogue and expert investigation into novel questions and issues, and to enable new ways to anticipate the future consequences of today’s decision.
Inclusive dialogue and debate
Of course an essential element to this governance landscape is an ongoing dialogue with wider society about how data is used. Our principles emphasise inclusivity in the management of dilemmas about data use, and building trust in data science and digital technologies requires engagement and debate.
As part of the British Academy and Royal Society study, a literature review of a selection of public engagement activities examined what we know about public dialogue on data use. To be effective, data governance needs to be grounded in engagement. Such engagement needs to be a dialogue rather than a one-way activity; be open; have the capacity to influence policy; and explicitly articulate competing values at stake.
The literature review showed that, as methodologies for the collection and analysis of data evolve rapidly, there is a need to revisit questions over time. Only very few studies investigated attitudes to new and future uses of data. There is also clearly a need for engagement with diverse communities. Studies highlighted that some groups within society could find it difficult to assess the benefits and risks of data uses, and that different generations have distinct relationships to data and its applications.
A foundation for trust
Initiatives such as Understanding Patient Data create the foundation for meaningful and valuable dialogue on the use of data, enabling a wider range of stakeholders to participate and engage in the ways their data is used. The stewardship body we recommend can help to ensure that such activities are embedded across the all sectors. At this moment of rapid change, taking time for real debate, and for the consideration of governance from a foundation of clear principles, will ensure that we build the trust necessary to access the significant benefits of better use of data.