Where next for proposals from India’s Committee of Experts on Non-Personal Data? In conversation with Astha Kapoor and Sylvie Delacroix

In the next of our series of ‘in conversation with…’ blogs, Sylvie Delacroix speaks to Astha Kapoor (Co-founder, Aapti Institute) about how discussions about data trusts are developing in India, following publication of a report from India’s Committee of Experts on Non-Personal Data. 

There has been a lot of interest in ideas from India’s Committee of Experts on Non-Personal Data about the development of data trusts. What is the thinking behind these ideas?

Astha: In India, like in many countries, the Government has been increasingly concerned about the idea of data sovereignty, and how to make sure that data generated by Indian citizens is governed and used for national benefit. The Committee looked at two main policy areas in this respect: how government can acquire data held by the private sector and increase data sharing for public value, and how to distribute the value created by data sharing across communities. 

The Committee’s focus on non-personal data has brought a different perspective to the conversation about data trusts. What are the implications of this?

Sylvie: A lot of our current legislation distinguishes between personal and non-personal data, creating extra safeguards around the use of data about individuals. But this is a remnant of an increasingly old-fashioned way of thinking about categories of data. The way we use data today means that anything that relates to human activities could be or become ‘personal’ data.  

Astha: Non-personal data is also a difficult concept – it is a definition by exclusion, and its interpretation is influenced by wider policy agendas around access to data. What we really need are new purpose-focused categories of data to guide how we govern data use. There is a role for regulators here to define for what purposes data sharing is appropriate or not – setting the contours of a data sharing ecosystem – and how practices like anonymisation can support data sharing for that purpose. 

Sylvie: It is too often the case that we think anonymisation is the answer to our data sharing concerns. It might be helpful in the short-term, but in the long-term we need to think about how we address the systemic weaknesses in the data governance landscape, in particular our ability to shift the asymmetries of power over data use that disenfranchise so many people. That question is bigger than anonymisation and requires different types of intervention. Chronology matters, and right now the personal / non-personal distinction is necessary because of how technologies and laws have evolved, but in the longer term, if our push for new data intermediaries is successful, there should be ways of redefining these debates.

Astha: Using ‘public value’ as a principle for data use could open some interesting gateways for data sharing, but there are big questions about how we would define public value, and the rights that individuals or communities should have over data that might have such value. 

Sylvie: Defining purpose and categories of value is difficult. Distinctions between what constitutes private and public value are always contested. As data intermediaries try and create new forms of value from data, there could be lessons to draw from the evolution of corporate social governance in relation to the need to overcome the public v. private dichotomy.

A lot of work on data trusts focusses on the rights that individuals have over data about them. Policy frameworks in India are developing ideas about community rights. What new issues or tensions do these raise?

Astha: The kernel of the Committee’s proposal around community data governance is that communities should have rights over the data they generate, and that an intermediary like a data trustee can help distribute the value created by the use of that data. This taps into a wider discussion about community data rights, drawing from experience that data acquires value when it is aggregated and from developing understandings of the relational aspects of data use. There is lots of research already around how we can conceptualise these rights, but also lots more work to do to nuance the ideas that are in the Committee’s report.  

Sylvie: Communities are not a single entity: they are fragmented, different members have different aspirations, which lead to conflict. The idea of having community rights is appealing in lots of ways, but focussing only on community rights risks marginalising the voices of minorities within that community. Show me a community where all want the same thing!

Astha: The power dynamics within communities are definitely something that we need to consider further, as well as their relationship with trusts that focus on specific domains. Healthcare, for example, was identified as a priority area by the Committee’s report. If we begin to develop community-focused data trusts, what are the implications for how sectorally-focused data intermediaries might operate? 

Sylvie: You cannot put the same right in multiple trusts. So it would seem like the way to have both sectoral and non-sectoral trusts would be to think about a stack of verticals – a community trust, then sector trusts that sit above them and that negotiate with the different community trusts. This approach would avoid having horizontal conflicts of rights between trusts, and avoid a situation in which individuals have to keep track of which data rights are being stewarded by which trust. The important question is then: what type of community trust would we envisage? Here the term ‘community’ may end up having to be an entirely voluntary, constructed one, which does not seek to track any particular social, economic or geographical identity (given the fact that we all have multiple, overlapping and constantly redefined identities). How can such trusts capture the fluidity of the views of their members?

Data trusts will develop in different ways in different jurisdictions, based on local needs and opportunities. Are there existing structures in India that could form the basis of data trusts in the future?

Astha: India has a robust cooperative movement, and there is a rich history of cooperatives that have grown out of trusted circles of people, often organised around a purpose, such as access to credit, milk production, or farming. It is possible to imagine a new data institution growing out of one of these existing bodies. A fisheries cooperative we are starting to work with, for example, brings together 50-100 members that support each other to access business loans. It provides banks with the financial records of the cooperative and data about productivity of the fisheries it represents, as evidence to show the loans can be repaid. As well as this type of cooperative, there are civil society organisations trying to empower communities, which could take on a trustee-like role. 

Sylvie: What makes data trusts special as a governance approach is that they combine collective action with independent stewardship and strong institutional safeguards. To what extent are these characteristics embedded in the type of cooperative you are talking about?

Astha: A lot of these cooperatives are very cohesive – the economic and social lives of their members are deeply intertwined. So the community itself has a lot of self-governing behaviours, either through cooperative membership rules or its patterns of voting. These organisations can also be held accountable through the courts, and there are instances of cooperatives being taken to court, especially for financial mismanagement. This isn’t perfect – relying on direct democracy within a cooperative risks marginalising minority voices, and there are also questions about the accessibility of justice. We need to think further about these social mechanisms for accountability, but it is exciting to see organisations emerging that could be a precursor to data trusts. 

Sylvie: This area of non-legal mechanisms for accountability seems like it would be interesting to explore further, to understand what behaviours or frameworks can help embed accountability in a way that complements formal governance interventions.

What is happening next with the proposal to develop community data trusts in India?

Astha: We are expecting a third draft of the committee of experts for public consultation. The report has a lot of unresolved questions still - – about whether government should be able to mandate data sharing; about public value; and about what type of rights we want communities to have over data – and we hope for a more consultative and prolonged process on these questions. 

 Author: Jess Montgomery (2021)