Robert Kneschke - Fotolia

Royal Society and British Academy call for data governance body

The Royal Society and the British Academy have jointly released a report urging the UK government to set up a public body to oversee data use

The Royal Society and the British Academy have joined forces to call for a new public body to oversee data use in UK society.

The report, Data management and use: governance in the 21st century, calls for an independent body to be set up to steward an “overall framework that can safeguard public confidence and ensure that the potential benefits of data use, such as improved public services, better healthcare and business innovation, are fully realised”.

Ottoline Leyser, chair of the Royal Society Science Policy Advisory Group and co-chair of the report, said today, on the Today programme on BBC Radio 4: “We are keen that the potential [of data science] is captured. At the moment, the main risks are in applications not being delivered because of the uncertainties in governance.

“The extent to which there is well-founded public trust in these systems [is an issue]. A good example is, where there was huge potential to be gained to from linking healthcare records in the NHS, but there was not enough open, clear, participatory discussion about how that system should work and so it stalled.”

She went on to question the loss of autonomy implied by algorithmically generated online shopping recommendations.

“I’m not sure I want a machine telling me what I should buy,” said Leyser. “I want the full range of options to decide for myself. And, more seriously, I want a person to explain why a diagnosis has been made [algorithmically, in a medical context].”

In a statement, she said: “We have reached a stage where most aspects of our day-to-day lives generate data that is collected, presenting opportunities for various actors to use this information. Many of the ways in which the data is used lead to positive impacts for us and wider society. However, the rapid rate of change in this area requires a new approach to governance that can keep pace, ensuring that the risks and benefits of new applications can be debated in a transparent and inclusive way.”

Data management proposals

The Royal Society and the British Academy brought together academics from the sciences, humanities and social sciences to review the current data management scene and write the report. They found that data usage, collection and management are “increasingly intertwined”, and “new ways of using data make it difficult to define which data is sensitive”.

The report’s authors – drawn from the “two cultures” of the arts and science in British academia, sometimes mutually antipathetic – found what they describe as “clear gaps and too many silos” in how data is used in the UK.

It proposes “an over-arching principle that systems that govern data need to promote human flourishing, focusing on ensuring that data is used to serve individuals and communities”.

The body whose creation the report recommends “would anticipate, monitor and evaluate the management and usage of data, build practices and set standards and provide clarity and propose solutions where tensions arise.

It would steward, rather than replace, a range of existing public and private actors, such as the Information Commissioner’s Office and those in the health sector. It would be led by experts from across disciplines including those who could represent the public interest. It would be UK focused but globally relevant and exchange good practice with other countries”.

The report does not, it said, “make specific recommendations about the location, funding or precise status of a new stewardship body for data governance”.

Genevra Richardson, vice-president for public policy at the British Academy and co-chair of the report, said: “The capturing and use of data is not a new concept – but the collection methods, usage and storage has changed significantly and continues to evolve. The National Academies have been able to convene the leading experts from the whole spectrum of disciplines and communities to look at these complex challenges. Our report presents a reasoned, evidence-based and pragmatic set of proposals.”

Antony Walker, deputy CEO of TechUK and member of the Royal Society and British Academy Working Group on Data Governance, added: “As we consider the implications of a world where humans live and work alongside intelligent machines we have to ensure that we put the interests of humans first and develop effective mechanisms to ensure we have full control over such machines.

Read more about data governance in civil society

“A data stewardship body, bringing together leading experts from academia, business and other fields, will be able to do just that, focusing, in particular, on promoting human flourishing to ensure that we anticipate and head off any risks that will arise from the development of AI [artificial intelligence] and harness the huge opportunities these advances will unlock.”

Data ethics council

In a similar vein, last year, the government accepted a House of Commons Science and Technology Committee’s recommendation that a data science ethics council be established at the Alan Turing Institute.

This was in response to a big data report published by the House of Commons select committee on science and technology in February 2016.

The committee’s chair, the then Conservative member of Parliament (MP) Nicola Blackwood, told a TechUK audience, in April 2016, that the committee’s recommendation that a data ethics council be established to oversee the government’s use of personal datasets had been accepted, and hoped to see “some progress on that”.

Blackwood narrowly lost her Oxford West and Abingdon seat in the recent snap general election.

In the same Today programme slot, in which Leyser was interviewed, government digital minister Matt Hancock welcomed the report.

“This is close to a proposal we have made for a data use and ethics commission,” he said. “We need to have strong governance some of which will be regulatory, but some of which will be about the norms of the businesses and the scientists that are making these [data science] innovations. This technology can be amazing, but it needs to be constrained so that the risks are mitigated.”

Read more on Healthcare and NHS IT