Already here, not yet everywhere: shaping the economic impact of artificial intelligence

Abstract Artificial intelligence (AI) is emerging as a general-purpose technology with far-reaching economic implications. Whilst its capabilities are already widely deployed across applications, such as text generation, coding, and forecasting, its aggregate impact on productivity and growth remains only partially visible—mirroring historical experiences with technologies like electricity. However, AI adoption is proceeding at unprecedented speed, suggesting potentially shorter diffusion lags and a more rapid translation into productivity gains. Estimates indicate that AI could significantly increase labor productivity growth in advanced economies, although outcomes remain uncertain and depend on the pace of adoption and complementary adjustments. The effects on inflation are ambiguous. On the one hand, efficiency gains and cost reductions may dampen price pressures; on the other, increased demand for investment, intermediate goods, and especially energy may exert upward pressure over time. At the same time, AI introduces new financial stability risks, including concentration amongst technology providers, more correlated decision-making, and heightened cyber threats. Europe’s position in the global AI landscape reflects both structural weaknesses—particularly in investment and infrastructure—and notable strengths in research and specialised industry. Fully harnessing AI’s potential requires not only innovation and capital but also effective institutional frameworks, reliable energy supply, and coordinated policy efforts.