Overview
As the DPI approach has increasingly been promoted and discussed as common sense at global levels, the realities, as we have shown here, are much more complex, ambiguous and misleading. This section provides a series of deep dives intended to provide a greater clarity on some current and emerging developments.
The first four deep dives (into key concepts, the evidence base, flows of money, and public rails for private innovation emphasis) highlight some of the productive dynamics that drive the DPI agenda. We aim to provide more clarity into exactly what is being promoted, how and by whom. The latter two deep dives, into DPI in Europe and DPI implications for AI conversations, focus on two recent developments in the DPI agenda. Again, the focus is understanding how we have gotten to where we are at with DPI, and emerging and likely future risks and dynamics.
Narrative building through the manipulation and ambiguity of key concepts
India’s G20 presidency provided a key moment at which the language around DPI seemed to achieve some consensus and credibility. The Digital Economy Working Group presented the definition:
“Under the Indian Presidency’s initiative, we recognise that digital public infrastructure, hereinafter referred to as DPI, is described as a set of shared digital systems that should be secure and interoperable, and can be built on open standards and specifications to deliver and provide equitable access to public and / or private services at societal scale and are governed by applicable legal frameworks and enabling rules to drive development, inclusion, innovation, trust, and competition and respect human rights and fundamental freedoms.”4
This definition has been repeated in other India G20 documentation (e.g. [here](https://g7g20-documents.org/fileadmin/G7G20_documents/2023/G20/India/Sherpa-Track/Digital Economy Ministers/2 Ministers'%20Annex/G20_Digital%20Economy%20Ministers%20Meeting_Annex1_19082023.pdf) and here). as well as more widely, e.g. WEF, UN SDG Digital Acceleration Agenda (2023).
Despite the G20 moment, defining DPI remains contestable and ambiguous. There is still not one circulating and clear definition for DPI. Different organisations and individuals defining DPI by essential characteristics, a set of parameters or even by its assumed outcomes.
Looking across some of the ways that DPI is presented by some of its key proponents (including the DPGA, CDPI, Co-Develop, Rockefeller, Gates Foundation, Carnegie Endowment), we identify and discuss four approaches that have been taken to defining DPI. These are used in combination and/or isolation.
DPI = a set of case studies
At its most specific, DPI is equated - at least in its first substantiation - with the three initial focus areas for India’s DPI: digital ID, payments and data exchange (e.g.
Gates, Co-Develop). Some definitions are broadening out from India’s experience to consider other areas where ‘DPI’ could be seen. CDPI categorises DPI cases into five areas: identities and registries, payments, data sharing and credentials, trust infrastructure, and discovery and fulfilment. Sometimes, definitions that start broad, defining DPI by a set of principles, end up quickly turning to concrete examples of ID, payments and data exchange to make concrete otherwise vague claims. The inconsistent case-study driven approach can be seen across Co-Develop’s “DPI exemplar stories”, UCL’s “DPI Map”, and the Indian G20’s “Global DPI Repository
(GDPIR)” – which all have different constructions and examples listed. For example, the GDPIR includes health, education, and agriculture as categories in addition to the ID, payments, data trio.
DPI = intended outcomes
Some definitions include an assumption that DPI will result in particular societal outcomes. There is little scope for questioning or interrogation of theory of change pathways here, as outcomes are part of what DPI is said to be. Examples include:5