Artificial Intelligence and Productivity Growth: From Firm-Level Efficiency to Macroeconomic Transformation

Abstract

Artificial intelligence (AI) is widely described as a general-purpose technology with the potential to alter productivity, employment, investment, and the distribution of economic gains. This paper synthesizes recent OECD and IMF evidence on the mechanisms through which AI can affect productivity and evaluates the conditions required for macroeconomic benefits to materialize. The analysis is based on comparative policy and literature synthesis rather than a new econometric estimation. Recent OECD modelling suggests that AI could add approximately 0.25–0.6 percentage points to annual total factor productivity growth and 0.4–0.9 percentage points to labour productivity growth over a ten-year horizon under specified adoption assumptions. At the same time, IMF evidence emphasizes that empirical estimates of AI’s employment and productivity effects remain inconclusive and that regulatory approaches differ substantially across countries. The paper develops a transmission framework linking AI capability to adoption, complementary investment, organizational redesign, skills, competition, and diffusion. The findings indicate that AI should not be treated as an automatic productivity shock: realized gains depend on complementary capital, managerial capacity, worker adaptation, data quality, and institutional conditions. The policy implication is that productivity-oriented AI strategies should combine innovation incentives with competition policy, skills development, worker transition support, and governance mechanisms that reduce concentration and unequal access.

Keywords

artificial intelligence; productivity; economic growth; labour markets; innovation

Authors & affiliations