THE ECONOMICS OF ARTIFICIAL INTELLIGENCE: PRODUCTIVITY, LABOUR MARKETS, AND STRUCTURAL CHANGE
Abstract
The rapid diffusion of generative artificial intelligence since 2022 has reignited long-standing debates in economics about the relationship between technological change, productivity growth, and labour-market outcomes. This article synthesises recent empirical and theoretical literature to assess how AI adoption is likely to affect aggregate productivity, occupational structure, and income distribution. Drawing on task-based models of automation and early firm-level evidence, the article argues that generative AI differs from earlier waves of automation in its capacity to substitute for cognitive, non-routine tasks previously considered resistant to computerisation, which broadens its potential impact to include highly educated, white-collar occupations. At the same time, measured productivity gains have so far been modest and unevenly distributed across firms and sectors, consistent with historical patterns of delayed diffusion for general-purpose technologies. The article concludes that policy attention should focus on complementary investment in human capital, competition policy to prevent excessive market concentration among AI providers, and social-insurance mechanisms capable of managing a potentially compressed transition period.