From Frugal Innovation to Sustainable Frugal AI: Scaling Cross-Border Inclusive Innovation

Abstract: Frugal innovation (FI) focuses on reducing costs, streamlining functionality, and optimising resources to tackle global issues of accessibility, affordability, and sustainability. Originated from developing economies, FI emphasises “make do”, high-impact solutions created under constraint. Extending these principles, Frugal AI is a novel approach in AI that leverages lightweight models, decentralised architectures, and open-source frameworks to achieve resource-efficient and functionally sufficient systems. By reducing computational and energy demands, it addresses the alarming challenges of the high resource intensity of conventional AI, supporting broader adoption in resource-constrained environments and lowering barriers for small businesses and underserved communities. However, little is known about how Frugal AI can be systematically developed, deployed, and scaled globally. This study conceptualises Frugal AI through development and deployment principles, identifying its core characteristics and mechanisms for impact. Drawing on secondary data from China, India and Vietnam, it proposes a baseline process model of Frugal AI development and deployment, tracing these countries’ evolution and illustrating how emerging economies pioneer scalable, inclusive AI solutions that address infrastructural and financial barriers. The study contributes to debates on reverse innovation, R&D internationalisation, and sustainable, democratised AI, and outlines a future research agenda for more equitable, resource-conscious global innovation.