Small AI as Digital Public Infrastructure: A New Paradigm for Resource-Constrained Economies
Keywords:
Artificial Intelligence, Edge AI, Digital Public Infrastructure, Inclusive development, Small AI, Sustainable developmentAbstract
Digital Public Infrastructure (DPI) has emerged as one of the key enablers for the design and implementation of development policies, facilitating scalable access to both public and private digital services. The current discussion regarding DPI is centered largely on three core components: digital identity systems, interoperable payment platforms and data exchange mechanisms. At the same time, rapid advances in artificial intelligence (AI) are creating demand for an additional infrastructural layer capable of supporting intelligent, adaptive and low-cost service delivery, particularly in resource-constrained environments. In response to this gap, this paper introduces Small AI as a new layer of DPI designed to advance inclusive and resilient development. Unlike frontier AI systems, which depend on high computational power, concentrated infrastructure and significant financial resources, Small AI systems are resource-efficient, locally adaptable and economically accessible. The paper conceptualizes Small AI not merely as a technological tool, but as a form of public infrastructure that delivers broad societal benefits in the same way as roads, electricity and telecommunications networks. Using a qualitative conceptual research design supported by comparative policy and institutional analysis, the study presents the Small AI–DPI Framework organized into three functional layers: Small AI compute capacity as a shared public utility, Small AI model repositories as open knowledge commons and Small AI application ecosystems as inclusive multi-sector service delivery platforms. Together, these layers illustrate the transformative potential of Small AI in enhancing institutional efficiency, financial inclusion, agricultural modernization and public service delivery. The paper further examines governance, financing, and scalability strategies for operationalizing Small AI across developing economies. It concludes with recommendations for multilateral development institutions, including the establishment of a Global Small AI Infrastructure Fund and a Small AI Readiness Index.