The Multiplicative Impact of AI Preparedness on E-commerce Growth: Empirical Evidence from the Global South Economies

Abstract

Artificial Intelligence (AI) is reshaping how economies transition toward digital services, with e-commerce emerging as a central driver of growth. Yet, countries with similar levels of AI readiness often show very different outcomes. This raises an important question: why does technological preparedness not translate into comparable economic gains across contexts?
This study examines that gap by introducing the Digital Buffer Hypothesis, which argues that the benefits of AI depend on a minimum level of internet infrastructure. Using cross-sectional data from 35 economies (2023–2024), an Ordinary Least Squares (OLS) model is employed to estimate the relationship between the e-commerce share of GDP and the IMF AI Preparedness Index. The model incorporates an interaction term for internet access, along with controls for GDP per capita and youth population share.
The results indicate that the model explains a substantial share of variation in digital structural change (Adjusted R² ≈ 0.49). AI preparedness emerges as a significant driver of e-commerce expansion, but its impact is notably stronger in countries with higher levels of connectivity. At the same time, basic internet access on its own does not appear sufficient to generate structural change, pointing to the importance of how infrastructure is utilized rather than its mere presence.
Taken together, the findings suggest that AI functions as a conditional driver of growth. Its economic returns depend on the presence of a supporting digital environment. For policymakers, particularly in the Global South, this implies that investments in AI capabilities need to be complemented by Digital Public Infrastructure that reduces friction and allows these technologies to scale effectively.