Harnessing Generative AI in Developing Economies: A Systems Framework for Policy Design in Bangladesh

Abstract

This thesis develops a systems-based policy framework for Generative Artificial Intelligence (GenAI) implementation in developing economies, with specific application to Bangladesh. While GenAI's potential productivity and labor market impacts are well-studied in developed economies, limited research addresses the challenges faced by developing countries positioned primarily as technology consumers rather than producers. The research employs causal loop diagramming to map interactions between five critical policy domains: human capital development, digital infrastructure, data sovereignty, sectoral stimulus, and governance.

The resulting framework identifies four primary reinforcing mechanisms that can accelerate adoption and three balancing mechanisms related to labor displacement. To validate the framework, the research analyzes contrasting implementation approaches from India and Egypt, demonstrating the importance of cross-domain synergies in effective policy design.

Applied to Bangladesh, the framework yields a dual-entry strategy focusing on healthcare and education sectors as initial implementation domains, leveraging the country's strategic advantages while addressing resource constraints through a consortia-based implementation model that creates institutional resilience. The thesis contributes both a reusable conceptual toolkit for analyzing GenAI policy in resource-constrained settings and an initial context-anchored roadmap for Bangladesh. Future research should refine the framework through longitudinal case studies while developing more detailed, stakeholder-engaged implementation plans for Bangladesh that include concrete budget allocations, institutional responsibilities, and measurable outcomes.