Experimental evidence on the productivity effects of generative artificial intelligence
Open accessThe DOI matches its Crossref record; 31 of the 31 references it cites were found in a registry or an index.
Abstract
We examined the productivity effects of a generative artificial intelligence (AI) technology, the assistive chatbot ChatGPT, in the context of midlevel professional writing tasks. In a preregistered online experiment, we assigned occupation-specific, incentivized writing tasks to 453 college-educated professionals and randomly exposed half of them to ChatGPT. Our results show that ChatGPT substantially raised productivity: The average time taken decreased by 40% and output quality rose by 18%. Inequality between workers decreased, and concern and excitement about AI temporarily rose. Workers exposed to ChatGPT during the experiment were 2 times as likely to report using it in their real job 2 weeks after the experiment and 1.6 times as likely 2 months after the experiment.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“That generative AI particularly benefited less able writers is paralleled in recent studies focusing on other domains in which generative AI has been shown to help less productive workers (4, 5).”
From Generative artificial intelligence enhances creativity but reduces the diversity of novel content (2023) · cites this paper as [4]
In the citing paper
“Previous studies have found that LLMs increase productivity the most for low-performing workers (Brynjolfsson et al. 2023, Noy and Zhang 2023, Dell'Acqua et al. 2023).”
From AI Meets the Classroom: When Do Large Language Models Harm Learning? (2024) · cites this paper as Noy and Zhang 2023
In the citing paper
“Noy and Zhang (2023) show that access to LLMs reduced completion time of mid-level professional writing tasks by 40 percent and raised quality by 18 percent, while compressing productivity differences between workers.”
From Labor Market Consequences of Generative AI: Early Evidence from Norway (2026) · cites this paper as Noy and Zhang, 2023
In the citing paper
“The study finds that AI can generate a large number of ideas within a very short time, markedly improving work efficiency (Noy and Zhang, 2023).”
From From humans to AI: understanding why AI is perceived as the preferred co-creation partner (2025) · cites this paper as Noy and Zhang, 2023
In the citing paper
“GenAI has signi cantly impacted various aspects of daily life, including enhancing productivity [68], writing [57, 58], interpersonal communication [35], and relationships [34, 76].”
From Creativity in the Age of AI: Evaluating the Impact of Generative AI on Design Outputs and Designers' Creative Thinking (2024) · cites this paper as [68]
33 more claims about this paper
Create a free account to see every claim other papers make about this one.
Citations
Cites
Robots and Jobs: Evidence from US Labor Markets
Acemoglu, D., & Restrepo, P. (2020) · Journal of Political Economy · doi:10.1086/705716 · Cited by 4,088
Why Are There Still So Many Jobs? The History and Future of Workplace Automation
Autor, D. H. (2015) · Journal of Economic Perspectives · doi:10.1257/jep.29.3.3 · Cited by 3,917 · cited 1×
“Previous waves of automation predominantly impacted “routine” tasks consisting of explicit sequences of steps that could be easily codified and programmed into a computer (Autor and Dorn, 2013; Autor, 2015).”
BackgroundThe growth of low-skill service jobs and the polarization of the US labor market
Autor, D. H., & Dorn, D. (2013) · American Economic Review · doi:10.1257/aer.103.5.1553 · Cited by 3,824 · cited 1×
“Previous waves of automation predominantly impacted “routine” tasks consisting of explicit sequences of steps that could be easily codified and programmed into a computer (Autor and Dorn, 2013; Autor, 2015).”
BackgroundAcemoglu, D., & Restrepo, P. (2018) · American Economic Review · doi:10.1257/aer.20160696 · Cited by 3,121 · cited 1×
“Moreover, their impacts on aggregate productivity may be small or nonexistent to the degree that they mainly serve to redistribute income previously earned by displaced workers to the capital owners supplying their robot replacements (10).”
BackgroundGPTs are GPTs: An Early Look at the Labor Market Impact Potential of Large Language Models
Eloundou, T., Manning, S., Mishkin, P., & Rock, D. (2023) · arXiv · doi:10.48550/arxiv.2303.10130 · Cited by 2,330
Data quality of platforms and panels for online behavioral research
Peer, E., Rothschild, D., Gordon, A., Evernden, Z., & Damer, E. (2021) · Behavior Research Methods · doi:10.3758/s13428-021-01694-3 · Cited by 1,718
Training, Wages, and Sample Selection: Estimating Sharp Bounds on Treatment Effects
LEE, D. S. (2009) · Review of Economic Studies · doi:10.1111/j.1467-937x.2009.00536.x · Cited by 1,399
Human Decisions and Machine Predictions
Kleinberg, J., Lakkaraju, H., Leskovec, J., Ludwig, J., & Mullainathan, S. (2018) · The Quarterly Journal of Economics · doi:10.1093/qje/qjx032 · Cited by 1,040
Brynjolfsson, E., Li, D., & Raymond, L. (2023) · National Bureau of Economic Research · doi:10.3386/w31161 · Cited by 995 · cited 1×
“These results are consistent with other studies showing productivity-enhancing and equalizing effects of recent AI technologies (8, 15, 16, 18).”
ResultToward understanding the impact of artificial intelligence on labor
Frank, M. R., Autor, D., Bessen, J. E., Brynjolfsson, E., Cebrian, M., Deming, D. J., Feldman, M., Groh, M., Lobo, J., Moro, E., et al. (2019) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.1900949116 · Cited by 713
Research works printed but not linked: 1
- Data and code for: S. Noy W. Zhang Experimental evidence on the productivity effects of generative artificial intelligence Open Science Framework (2023); https://osf.io/xd7qw/?view_only=b8dac58dd6f44b979bf81069022fa392DOI10.17605/OSF.IO/XD7QW.10.17605/OSF.IO/XD7QW · doi:10.17605/osf.io/xd7qw
Cited by
Items 1,801–1,859 of 1,859
Flugelman, M. Y. (2026) · Teaching and Learning in Medicine · doi:10.1080/10401334.2026.2737803
Chai, Y. (2026) · Digital Transformation and Society · doi:10.1108/dts-01-2026-0039
Used but Unseen: Generative AI Diffusion and the Governance Readiness Gap in Croatia
Momčilović, A., Rogge, N. D., & Vlahov, R. D. (2026) · Technology and Society · doi:10.1016/j.techsoc.2026.103567
Сердюк, О. (2026) · Economy of Industry · doi:10.15407/econindustry2026.03.023
Glebova, A. G. (2026) · Theoretical economics · doi:10.52957/2221-3260-2026-8-126-139
Meng, X. (2026) · Asian Economic Policy Review · doi:10.1111/aepr.70059
Auditing behavioral claims about AI: Two public-data applications
Tang, K. (2026) · Social Sciences & Humanities Open · doi:10.1016/j.ssaho.2026.103760
Zhao, K., Zhang, B., & Qiao, X. (2026) · Asian-Pacific Economic Literature · doi:10.1111/apel.70070
Li, Z., Liu, H., Shan, S., Zhang, M., & Yang, Q. (2026) · Internet Research · doi:10.1108/intr-10-2025-1768
Ilomuanya, S. E., Oladosu, M. A., MUSTAPHA, T. O., & Abah, M. A. (2026) · International Journal of Soft Computing and Engineering · doi:10.35940/ijsce.e3728.16051126
CHIU, Y.-H., & Chen, C.-C. (2026) · Behaviour and Information Technology · doi:10.1080/0144929x.2026.2732025
Yousef, M., van Hierden, Y., Dietrich, T., Zaman, K., & Saleme, P. (2026) · Marketing Intelligence & Planning · doi:10.1108/mip-03-2025-0271 · Cited by 1
Berger, S., & Voigt, J. (2026) · Creativity and Innovation Management · doi:10.1111/caim.70090 · Cited by 4
Firms Diverged Faster than Households: Market-Capitalization Concentration in the S&P 500
Pattnaik, S., Jain, C., & Pinsky, E. (2026) · Journal of risk and financial management · doi:10.3390/jrfm19100747
Сухинов, А. И., Kharitonov, K. A., & Ugnich, E. A. (2026) · Advanced Engineering Research · doi:10.23947/2687-1653-2026-26-3-2796
How Artificial Intelligence Shapes Human Narratives
Priniski, J. H. (2026) · Journal of Language and Social Psychology · doi:10.1177/0261927x261493082
Suppressed inquiry: The hidden toll of working with satisfaction-optimized AI
Carmon, Z., Linzen, I., Sela, A., & Steinhart, Y. (2026) · Current Opinion in Psychology · doi:10.1016/j.copsyc.2026.102453
Shukeyev, U. (2026) · Bangladesh Journal of Multidisciplinary Scientific Research · doi:10.46281/bjmsr.v11i6.2959
Ekmektzoglou, K. A., & Rokkas, T. (2026) · Frontiers in Artificial Intelligence · doi:10.3389/frai.2026.1943025
Liu, J. (2026) · Journal of risk and financial management · doi:10.3390/jrfm19100755
Zhang, X., Fang, J., Chen, S., & Luo, J. (2026) · JMIR Nursing · doi:10.2196/100916
“These observations converge with recent qualitative scholarship, highlighting AI’s capacity to alleviate administrative burdens and optimize workflow management [6,21,26].”
Liu, J., Zhong, C., Meng, F., Shank, D. B., & Wunsch, D. C. (2026) · International Journal of Information Management · doi:10.1016/j.ijinfomgt.2026.103140
The circular economy, explained by AI
Morseletto, P., & Bocken, N. M. P. (2026) · Journal of Industrial Ecology · doi:10.1007/s44498-026-00192-z
Jongkind, R. C., van Berlo, B., Otto, S., et al. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-37982-5_19
Artificial Intelligence and the Bioeconomy: Implications for the Annual Review of Resource Economics
Rausser, G. C., & Zilberman, D. (2026) · Annual Review of Resource Economics · doi:10.1146/annurev-resource-060126-044150
Too Much Mind Extension? Towards a Virtue Ethics of the Extended Mind
Chlebowski, S., Ratajczyk, D., & Wyrwa, M. (2026) · Roczniki Filozoficzne · doi:10.18290/rf26743.6
The AI Productivity Paradox: When Readiness and Knowledge Sharing Meet Anxiety
Popatia, M., Akhtar, S., Sameeni, M. S., & Karim, S. (2026) · journal of social sciences review · doi:10.62843/jssr.v6i4.769
Predicting the long-term impact of AI assistants on collective intelligencePreprint
Kim, J., & Park, S. W. (2026) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.64898/2026.09.29.755292
ЗАНЯТОСТЬ В МОСКВЕ В КОНТЕКСТЕ РАЗВИТИЯ ИСКУССТВЕННОГО ИНТЕЛЛЕКТА: ОТРАСЛЕВОЙ АНАЛИЗ
Белоновская, А. М. (2026) · Прогрессивная экономика · doi:10.54861/27131211_2026_9_508
Ethical Aspects of Generative Artificial Intelligence—A Comprehensive Scoping Review
Maini, E., Maini, B., Marwaha, D., & Kumar, S. A. (2026) · Lecture notes in electrical engineering · doi:10.1007/978-981-95-7449-0_30
Balancing structural life and digital technologies: global evidence for a preventive policy model
Özdilek, Ü. (2026) · Health Policy and Technology · doi:10.1016/j.hlpt.2026.101334
Editing by prompt. How professional editors use ChatGPT when revising business texts
Janssen, D., & IJlst, D. (2026) · Computers & composition/Computers and composition · doi:10.1016/j.compcom.2026.103038
Myers, R. W., Ramachandran, K., & Varadarajan, B. (2026) · INFORMS Transactions on Education · doi:10.1287/ited.2025.0177
Pin, L., & Huang, Z. (2026) · Economic Change and Restructuring · doi:10.1007/s10644-026-10081-x · Cited by 1
Fink‐Lamotte, J., Günther, F., Haaf, J. M., Specht, J., Cludius, B., Hilbert, K., & Saulin, A. (2026) · Psychologische Rundschau · doi:10.1026/0033-3042/a000788
Zhang, H., Wu, R., Hou, W., Lan, P., & Hu, S. (2026) · Data Science and Management · doi:10.1016/j.dsm.2026.100205
AI Diffusion and Short-Run Economic Performance: High-Frequency Evidence from U.S. BusinessesPreprint
Rabbani, M., & Nishi, S. F. (2026) · SSRN Electronic Journal · doi:10.2139/ssrn.7571286
Artificial Intelligence in Decision-Making: Building Future-Proof Strategies for Business Innovation
Efstratiadou, S.-M., & Lytras, M. D. (2026) · Emerald Group Publishing Limited eBooks · doi:10.1108/978-1-83662-878-120261011
Envidia: imitar, suplantar y aparentar
Forero, F. E. M. (2026) · doi:10.62486/978-9915-9928-6-0.ch04
The access–conversion gap across countries: how generative AI adoption exceeds learning use
Achuthan, K. (2026) · Frontiers in Education · doi:10.3389/feduc.2026.1946943 · Cited by 1
Серединская, Д. (2026) · Regional and Branch Economy · doi:10.47576/2949-1916.2026.9.9.027
Educación a lo largo de la vida: Aprender, desaprender y volver a aprender
Cardozo, G. T., & Gamboa, A. J. P. (2026) · doi:10.62486/978-9915-9928-4-6.ch15
Gula: consumir recursos y producir contenido sin medida
Navia, J. M. A., Zapata, E. M. J., & Castillo, V. S. (2026) · doi:10.62486/978-9915-9928-6-0.ch05
Tools of the trade: AI and the restructuring of the experience premium
Mattingly, E. S., Hyde, S. J., & Manikas, A. S. (2026) · International Journal of Manpower · doi:10.1108/ijm-04-2026-0378
Soberbia: creer que la inteligencia artificial lo sabe todo
de la Cruz, J. E. H., Polanía, L. R., & Castillo, V. S. (2026) · doi:10.62486/978-9915-9928-6-0.ch01
The Epistemic Leader: Developing Epistemic Evaluation Capability in AI Leadership
Nguyen, V. M. H., Attanayake, C., & Le, Q. K. (2026) · Advances in computational intelligence and robotics book series · doi:10.4018/979-8-2600-1488-2.ch011
Pereza: delegar aquello que deberíamos aprender, pensar o decidir
Cano, C. A. G., & Pérez, J. E. P. (2026) · doi:10.62486/978-9915-9928-6-0.ch07
Avaricia: concentrar el poder, los datos y los beneficios
Suárez, Y. S., Castillo, V. S., & Morales, C. E. B. (2026) · doi:10.62486/978-9915-9928-6-0.ch02
Lujuria: la fascinación desbordada por lo artificial
Gamboa, A. J. P. (2026) · doi:10.62486/978-9915-9928-6-0.ch03
Labor Market Consequences of Generative AI: Early Evidence from Norway
Facius, D., & Iacono, R. (2026) · CESifo Working Papers · doi:10.65864/8a4auliirt
Experimental Evidence on the Learning Impact of Generative AI
Contractor, Z., & Reyes, G. (2026) · arXiv · doi:10.48550/arxiv.2607.08849 · Cited by 10
Game Plot Design With an LLM-Powered Assistant: An Empirical Study With Game Designers
Alavi, S. H., Xu, W., Jojic, N., Kennett, D., Ng, R. T., Rao, S., Zhang, H., Dolan, B., & Shwartz, V. (2026) · IEEE Transactions on Games · doi:10.1109/tg.2026.3663566
Autor, D., Rodchenko, T., Martin, J., Iscenko, Z., Strand, S., Pearl, D., & Ferere, M. (2026) · doi:10.3386/w35720
Sun, C., & Yang, J. (2026) · The Manchester School · doi:10.1111/manc.70080
Generative AI impacts on intra-urban inequality and skill premium in Beijing
He, X., Zhao, H., Ma, M., Lai, E. W. C., Enomoto, K., Hu, A., Li, J., Chu, L., & Lai, Y. (2026) · arXiv · doi:10.48550/arxiv.2605.25505
The Broken Ladder: AI, Remote Work, and Early-Career Hiring
Lambert, P. J., & Schindler, Y. (2026) · doi:10.2139/ssrn.6787638
Performance and Metacognition Disconnect when Reasoning in Human-AI Interaction
Fernandes, D., Villa, S., Nicholls, S., Haavisto, O., Buschek, D., Schmidt, A., Kosch, T., Shen, C., & Welsch, R. (2024) · arXiv · doi:10.48550/arxiv.2409.16708 · Cited by 71
Metadata
- Authors
- Noy, S., & Zhang, W.
- Affiliations
- Massachusetts Institute of Technology (US)
- Corresponding author
- Noy, Shakked
- Venue
- Science
- Published
- 2023-07-14
- Volume
- 381
- Issue
- 6654
- Pages
- 187-192
- ISSN
- 0036-8075 (ISSN-L)
- 1095-9203
- OpenAlex
- W4384154918
- Semantic Scholar
- 8d020275181c69e5e768c6ffc40e09710a6f54f1
- Cited by
- 2,150
- References
- 29
- Citations by year
- 2026: 985 · 2025: 639 · 2024: 405 · 2023: 116
- FWCI
- 435.72
- Type
- journal article
- Language
- en
- Open access
- green
- doi.org
- Locations
- Science (publishedVersion)
- SSRN Electronic Journal (acceptedVersion)
- PubMed (publishedVersion)
- Topics
- Social Sciences › Economics, Econometrics and Finance › Economics and Econometrics › Labor market dynamics and wage inequality
- Physical Sciences › Computer Science › Artificial Intelligence › Artificial Intelligence Applications
- Physical Sciences › Computer Science › Artificial Intelligence › Ethics and Social Impacts of AI
- Keywords
- generative AI, ChatGPT, productivity, professional writing, artificial intelligence adoption, output quality, task completion time, attitudes toward AI, workplace technology use
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.