Tag Archives | large language models

Is it Okay to Use Generative AI in Sociological Research?

Oscar Stuhler, Dustin S. Stoltz, AJ Alvero, Marshall A. Taylor

Sociological Science November 17, 2026
10.15195/v13.a52


This article presents the results of a survey of what uses of generative AI (GenAI) sociologists consider acceptable across different aspects of the research process, including literature review and ideation, data collection and analysis, writing, and peer review. Despite some areas of agreement, sociologists diverge considerably in their views on the acceptability of GenAI use in research. Many of the 19 practices we asked about were deemed generally unacceptable by a significant share of respondents even when disclosed in resulting publications, while an often similarly sized share deemed them acceptable even without disclosure. For example, 40 percent consider it unacceptable to generate an outline for an article with GenAI while for 34 percent doing so is acceptable and need not be disclosed. Few sociologists categorically reject the use of GenAI in research, with only 5 percent finding all uses unacceptable. GenAI use is more acceptable to quantitative researchers, non-native English speakers, and men. Although 82.2 percent of sociologists support journals’ adoption of mandatory disclosure statements on GenAI use, only 22 percent trust such statements.
We reflect on these findings and discuss their broader implications.

Creative Commons LicenseThis work is licensed under a Creative Commons Attribution 4.0 International License.


Oscar Stuhler: Department of Sociology, Northwestern University.
E-mail: oms@northwestern.edu

Dustin S. Stoltz: Department of Sociology and Anthropology, Lehigh University.
E-mail: dss219@lehigh.edu

AJ Alvero: Center for Data Science for Enterprise and Society, Cornell University.
E-mail: aja326@cornell.edu

Marshall A. Taylor: CREST, ENSAE, Institut Polytechnique de Paris.
E-mail: marshall.taylor@ensae.fr

Acknowledgments: First and foremost, we want to thank all respondents of our survey who were willing to take the time to share their thoughts. This article exists because of their generosity. For feedback on survey design and prior drafts, we are indebted to Zackary Dunivin, Jan Fuhse, Patrick Kaminski, Ryan Leung, Santiago J. Molina, Laura K. Nelson, Étienne Ollion, Christine Percheski, and Cat Dang Ton. We are also thankful for comments received when presenting this work at the MZES speaker series at Mannheim University and the session on Implications of AI at the 2026 ASA Annual Meeting.

Corresponding author: Oscar Stuhler, E-mail: oms@northwestern.edu


Supplemental Materials

Reproducibility Package: A replication repository for this article can be found at https://osf.io/fpjzn. The data for this article are hosted on the Harvard Dataverse at https://doi.org/10.7910/DVN/HNK0SO.


  • Citation: Stuhler, Oscar, Dustin S. Stoltz, AJ Alvero, and Marshall A. Taylor. 2026. “Is it Okay to Use Generative AI in Sociological Research?” Sociological Science 13: 1382-1403.
  • Received: August 21, 2026
  • Accepted: September 29, 2026
  • Editors: Stephen Vaisey, Jeremy Freese
  • DOI: 10.15195/v13.a52


0
SiteLock