AI Diffusion to Low-Middle Income Countries; A Blessing or a Curse?

Rapid advancements in AI have sparked significant research into its impacts on productivity and labor, which can be profoundly positive or negative. Often overlooked in this debate is understanding of how AI technologies spread across and within economies. Equally ignored are developing economies facing substantial labor market impacts from rapid, and a loss in competitiveness, from slow AI diffusion. This paper reviews literature on technology diffusion and proposes a three-way framework for understanding AI diffusion: global value chains, research collaboration, and inter-firm knowledge transfers. This is used to measure AI diffusion in sixteen low-middle-income, and four developed economies, as well as to evaluate dependence on China and the USA for access to AI technologies. The study finds a significant gap in diffusion rates between the two groups, but current trends indicate it is narrowing. China is identified as a crucial future source of AI diffusion through value chains, while the USA is more influential in research and knowledge transfers. The paper's limitations include the omission of additional data sources and countries, and the lack of investigation into the relationship between diffusion and technology intensity. Nonetheless, it raises salient macro-level questions about AI diffusion and suggests emphasis on redistribution mechanisms of AI induced economic gains, and bilateral agreements as a complement to international accords, to address diverse needs and corresponding risks faced by economies transitioning into an AI-dominated era. Additionally, it highlights the need for research into the links between AI diffusion, technology intensity, and productivity; case studies combined with targeted policy recommendations; more accurate methods for measuring AI diffusion; and a deeper investigation into its labor market impacts particular to LMICs.