Generative AI

Authors:     Alex Bick, Adam Blandin, David Deming

Latest Draft:     10/2025 (St. Louis Fed Blog Post)

Generative AI Adoption Tracker.

Summaries:  VoxEU, The Project on Workforce

Press / Policy: Reuters, NPR: Planet Money,  Psychology Today, Speech by Fed Governor Lisa Cook , Harvard Gazette, The Economist, CNET, International Energy Agency (IEA), Marketplace, Barron's (feature), Barron's, Congressional Budget Office, Barron's, Inter-Governmental AI Safety Report, Bloomberg, Bloomberg, Slow Boring, Yahoo! Finance, Forbes, Forbes, Wall Street Journal, New York Times, Reason, Washington Post, AEI, Financial Times, The Economist, The Hill, Speech by Fed Governor Michael Barr, 2026 Economic Report of the President (CEA)

Abstract:     Generative artificial intelligence (AI) is a potentially important new technology, but its impact on the economy depends on the speed and intensity of adoption. This paper reports results from a series of nationally representative U.S. surveys of generative AI use at work and at home. As of late 2024, nearly 40% of the U.S. population age 18-64 uses generative AI. Among employed respondents, 23% used generative AI for work at least once in the previous week: 9% used it every workday, and 14% on some but not all workdays. Relative to each technology's first mass-market product launch, work adoption of generative AI has been as fast as the personal computer (PC), and overall adoption has been faster than either PCs or the internet. Generative AI and PCs have very similar early work adoption patterns by education, occupation, and other characteristics. Between 1 and 5% of all work hours are currently assisted by generative AI, and respondents report time savings equivalent to 1.4% of total work hours. This suggests that substantial productivity gains from generative AI are possible.

Authors:     Alex Bick, Adam Blandin, David Deming, Nicola Fuchs-Schundeln, Jonas Jessen

Latest Draft:     03/2026 (St. Louis Fed Blog Post)

Generative AI Adoption Tracker.

Press: Silicon Continent (Luis Garicano substack) 

Abstract:     This paper combines international evidence from worker and firm surveys conducted in 2025 and 2026 to document large gaps in AI adoption, both between the US and Europe and across European countries. Cross-country differences in worker demographics and firm composition account for an important share of these gaps. AI adoption, within and across countries, is also closely linked to firm personnel management practices and whether firms actively encourage AI use by workers. Micro-level evidence suggests that AI generates meaningful time savings for many workers. At the macro level, in recent years industries with higher AI adoption rates have experienced faster productivity growth. While we do not establish causality, this relationship is statistically significant and similar in magnitude in Europe and the US. We do not find clear evidence that industry-level AI adoption is associated with employment changes. We discuss limitations of existing data and outline priorities for future data collection to better assess the productivity and labor market effects of AI.

Authors:     Alex Bick, Adam Blandin, David Deming, and Tyler Schumacher

Latest Draft:     04/2026

Generative AI Adoption Tracker.

Abstract:     We measure how workers use genAI in their jobs using a nationally representative survey that links genAI use to detailed tasks. GenAI currently assists a broad range of work, with at least one in five workers using genAI in 80\% of occupations and 40\% of job tasks. Yet in most of these cases adoption rates remain below 50\%, with some individuals systematically adopting genAI for more tasks than others who perform similar work. As a result, although genAI "exposure" measures correlate positively with adoption, they explain only about half of the variation across workers. The nature of genAI use also varies across occupations, with workers in some occupations using it mainly for high-expertise work and others for low-expertise work. Survey-based task shares differ systematically from estimates from genAI platform chat-log data, primarily because chat data overclassify basic, generic tasks relative to the ONET framework. We organize our findings into occupation- and task-level adoption indexes, providing an input for research on the labor market effects of genAI.