The Illusionary Model of Relative Economic Growth in the Era of Artificial Intelligence

Artificial Intelligence (AI) is increasingly viewed as a major driver of productivity growth and economic transformation. This paper challenges the assumption that AI-related technological expansion necessarily generates widespread macroeconomic productivity gains. Drawing on endogenous growth theory and the productivity paradox literature, it introduces the Relative Economic Growth Illusion (REGI) framework, which argues that contemporary AI-driven growth may operate through concentrated and intangible-intensive economic structures rather than through broad productivity diffusion across the economy. Using cross-country evidence from OECD Productivity, OECD STAN, OECD Patents, INTAN-Invest, and Functional Urban Areas (FUAs) databases, the study combines descriptive analysis with panel and robust regression techniques to examine the relationship between AI innovation, productivity, and intangible capital accumulation. The results show that AI patent intensity has weak and statistically insignificant associations with aggregate Total Factor Productivity (TFP), while intangible capital accumulation remains strongly linked to localized sectoral productivity gains. Evidence also reveals a marked geographical concentration of AI-related innovation within a limited number of technologically advanced economies. These findings suggest that AI-driven technological progress generates localized efficiency improvements while diffusing only weakly across the broader economy. As a result, observed economic expansion may increasingly reflect concentrated growth driven by intangible capital and technological concentration rather than broad-based productivity improvements. JEL classification numbers: O33, O47, D43.