By Liza Paudel, Omidyar Network and Christopher Wilson, Mydata Global
With mobile payments; facial recognition in airports, public schools, and city streets; algorithms used in hiring decisions and social services provision, our lives are increasingly, and often invisibly, ‘datafied.’ We are entering 2024 with the discomfort that today data is collected with little agency on the part of consumers and concentrated with a handful of players, often behind proprietary walls, massively concentrating power and limiting outcomes that serve the public interest.
The mainstream introduction and diffusion of generative AI systems through platforms like ChatGPT has only exacerbated the problem and highlights an unequal AI future that lies ahead if the lopsided bargain of our data status quo continues unchecked.
Effective data governance at scale has thus never been more urgent, and there are important efforts being made towards the same. Work towards the UN’s Global Digital Compact — expected to create shared principles for our collective digital futures at the global level — is already underway; and after six months of negotiations, the EU AI Act passed last week. Yet, such initiatives remain hindered by general shortcomings in how we understand data’s role in society, how we anticipate the scope and limitations of the emerging data economy and the wider data ecosystem, and the ways in which we frame these issues in the context of policy and our everyday lives.
In this blog post, we reflect on what we have learned from recent events focused on data and data governance in 2023 to explore how a more nuanced discussion of these issues can help us to pursue a fairer data future. We draw on conversations at the Mydata Conference 2023, RightsCon 2023, the Council for a Fair Data Future, UNCTAD eWeek, and others, to suggest two fundamental challenges to how we understand the role of data in society and offer visions for moving forward, explored in-depth in two collaborative think pieces by our partners.
Our data discourse is not sufficiently nuanced or keeping up to speed.
The analogies we use to discuss data have been limited at best and unhelpful at worst, with the framing of data as property (e.g. “data as oil”) still the most prominent. At the Mydata Conference 2023 held in Helsinki in May, Maria Rautavirta from the Finland Ministry of Transport and Communications likened the way in which different stakeholders talk about data to a blind man describing an elephant — none incorrect per se, and none comprehensively correct either.
With rapid evolutions in technology, often both complex and opaque, the value and role of data also continues evolving, defying easy definitions. Take for instance the urgent, but open questions around how generative AI may alter or even redefine data governance — from individual data rights to intellectual property. At the same time, the urgency of popular discourse surrounding these novel developments often invites simplistic arguments that are not always nuanced enough to help manage the complex questions at play.
There is an over-reliance on privacy and consent and on the individual framing of data at the expense of collective frameworks and structural concentrations of power.
Most existing regulatory frameworks for data governance, such as GDPR, are premised upon the foundations of individual-level privacy protections and notice-and-consent. This represents a necessary but insufficient framework for both regulating and enabling data sharing at scale. Nor is it adequate to empower and protect individuals and consumers from the byzantine and often invisible business models of data extraction that dominate the data ecosystem today.
The arbitrary nature of the distinction between personal and non-personal data in the age of big data and AI systems is increasingly recognized, but that understanding has yet to lead to increased control and agency in the hands of consumers in the governance mechanisms that exist. Finally, the surrounding system-level constraints of capitalism and culture, and how they incentivize certain behaviors over others — the degradation of antitrust, social contract, and even democracy, for instance — also remain often unacknowledged.
With these challenges in mind, we are launching two think pieces with our partners at Mydata Global as a preliminary effort to advance our thinking:
First, In This Together: Combining Individual and Collective Strategies to Confront Data Power is a short paper exploring notions of individual data rights and collective data governance, and how the apparent tensions and ambiguities between the two can be resolved to enhance advocacy for better data governance. A collaborative effort between four Omidyar Network grantees — data governance experts and advocacy groups, Mydata Global,Aapti Institute, The Datasphere Initiative, and Connected by Data — this piece focuses on how these dynamics are manifest and suggests that advocacy efforts can benefit by leveraging the many overlaps and additive opportunities between these conceptual frameworks.
Second, Towards Network-based Ecosystems is a short technical paper exploring the implications of recent developments in customer data and relationship management, and proposing a network-based approach to empower individuals in their online interactions with vendors. It considers how recent trends in online consumer tracking, digital marketing, wallets, and loyalty platforms can be leveraged to usher in a paradigm shift in how customer relationships are managed online. It suggests a network-based infrastructure as the most promising to enable a fairer and more equitable data economy, and outlines design principles through which this could be pursued.
These are two modest and preliminary contributions to broader community efforts that are already underway to shift the current data paradigm.
As we look to 2024, we close 2023 with a changing data ecosystem, and a changed lexicon. “GPT,” “large language model (LLM),” and “generative artificial intelligence (GenAI)” have made their way into the dictionaryalongside the cultural mainstream. “AI” is the Word of the Year. What do all these novel advancements have in common? The need for more and more data — our data. As the role of data in our lives only grows bigger and more complex, there remain a multitude of obstacles to a fairer and more equitable data ecosystem. It does not help, however, to exaggerate misalignments or paint data governance as an overly complicated issue that experts and practitioners cannot agree on. Doing so only reinforces the status quo, distracting from opportunities for fresh thinking around better oversight and increased value in the public interest.
With these two think pieces, our partners contribute a combination of imagination and expertise to pursue a more expansively and equitably governed data future, helping advance our imagination of what is possible and giving us concrete tools to work towards that vision. Alongside them, Omidyar Network is committed to collaborative visioning and action to secure a data ecosystem that empowers individuals and communities and promotes trust and accountability to responsibly unlock the value of data for all. As noted in the title of one of these contributions, we’re all in this together.