From Digital Public Infrastructure to AI Governance: Lessons from India for the Global South

Digital Public Infrastructure, or DPI, has become the model that governments across the Global South are told to copy. India's Aadhaar identity system and its Unified Payments Interface sit at the center of that story. Now that the same governments are confronting artificial intelligence, a second claim has attached itself to the first: that whatever institutional strength built India's DPI should also help a state govern AI. This paper puts that claim to the test instead of taking it on faith. It compares India, Brazil, Indonesia and Kenya using a most-differentsystems design, tracing six candidate mechanisms through which DPI experience might generate AI governance capacity, to ask where that capacity actually comes from. Three rival hypotheses are weighed against the record: direct institutional transfer, technological discontinuity, and conditional adaptation. The evidence favours the third. DPI-derived capacity does carry over, but only in part, and only once it meets independent regulators, enforceable data-protection law and accountability mechanisms that DPI programmes do not generate on their own. To keep the claim testable, the paper treats AI governance capacity as a disaggregated construct covering regulatory, enforcement, technical, coordination, accountability, data-governance and compute capacity, and builds a DPI-to-AI Governance Framework that specifies when transfer works and when it does not. It ends with what this means for governments drawn to the "India model," and with an agenda for the comparative and quantitative work that should follow.