What ChatGPT Tells Us about Gender: A Cautionary Tale about Performativity and Gender Biases in AI
Open accessThe DOI matches its Crossref record; 30 of the 32 references it cites were found in a registry or an index.
Abstract
Large language models and generative AI, such as ChatGPT, have gained influence over people’s personal lives and work since their launch, and are expected to scale even further. While the promises of generative artificial intelligence are compelling, this technology harbors significant biases, including those related to gender. Gender biases create patterns of behavior and stereotypes that put women, men and gender-diverse people at a disadvantage. Gender inequalities and injustices affect society as a whole. As a social practice, gendering is achieved through the repeated citation of rituals, expectations and norms. Shared understandings are often captured in scripts, including those emerging in and from generative AI, which means that gendered views and gender biases get grafted back into social, political and economic life. This paper’s central argument is that large language models work performatively, which means that they perpetuate and perhaps even amplify old and non-inclusive understandings of gender. Examples from ChatGPT are used here to illustrate some gender biases in AI. However, this paper also puts forward that AI can work to mitigate biases and act to ‘undo gender’.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
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?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Recent research has shown that commercially available LLMs can perpetuate race-based medical misconceptions [3, 5, 6].”
From Evaluating and addressing demographic disparities in medical large language models: a systematic review (2025) · cites this paper as [3]
Experimental — accuracy not yet measuredThe cited source says what the claim attributes to it.
In the citing paper
“These models, however, are trained on human-generated text, which often contains biases [3–5].”
From Evaluating and addressing demographic disparities in medical large language models: a systematic review (2025) · cites this paper as [3]
The cited source says what the claim attributes to it.
In the citing paper
“On one hand, we must be mindful that several studies have underscored the ways in which the outputs of GenAI tools like ChatGPT can inadvertently reinforce biases in areas such as gender and politics (Gross, 2023; Motoki et al., 2024; Rutinowski et al., 2024).”
From If ChatGPT can do it, where is my creativity? generative AI boosts performance but diminishes experience in creative writing (2025) · cites this paper as Gross, 2023
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
On the Dangers of Stochastic Parrots: Can Language Models Be Too Big? 🦜
Bender, E. M., Gebru, T., McMillan-Major, A., & Shmitchell, S. (2021) · Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency · doi:10.1145/3442188.3445922 · Cited by 6,988 · cited 14×
“Bender et al. (2021) highlight the unfathomable amount of data that go into LLMs, and these data inevitably include stereotypical associations and derogatory portrayals of gender.”
BackgroundMethodInfluentialButler, J. (2004) · doi:10.4324/9780203499627 · Cited by 5,556
The Psychic Life of Power: Theories in Subjection
Butler, J. (1997) · doi:10.1515/9781503616295 · Cited by 4,949
Floridi, L., Cowls, J., Beltrametti, M., Chatila, R., Chazerand, P., Dignum, V., Luetge, C., Madelin, R., Pagallo, U., Rossi, F., et al. (2018) · Minds and Machines · doi:10.1007/s11023-018-9482-5 · Cited by 4,774 · cited 7×
“…way that allows for human self-realization without deval-uing human abilities; enhances human agency without removing human responsibility; increases social capabilities without reducing human control; and cultivates societal cohesion without eroding human self-determination (Floridi et al. 2018).”
BackgroundInfluentialPerformative Acts and Gender Constitution: An Essay in Phenomenology and Feminist Theory
Butler, J. (1988) · Theatre Journal · doi:10.2307/3207893 · Cited by 4,116
Algorithms of Oppression: How Search Engines Reinforce Racism
Noble, S. U. (2020) · doi:10.2307/j.ctt1pwt9w5 · Cited by 2,709
Ray, P. P. (2023) · Internet of Things and Cyber-Physical Systems · doi:10.1016/j.iotcps.2023.04.003 · Cited by 2,463 · cited 8×
“Given how the data for LLMs are collected and where these data come from, LLMs are biased by default (Bender et al. 2021; Ferrera 2023; Ray 2023).”
BackgroundMethodInfluentialFraser, N. (2000) · New Left Review · doi:10.64590/pup · Cited by 1,298
Gender stereotypes about intellectual ability emerge early and influence children's interests
Bian, L., Leslie, S.-J., & Cimpian, A. (2017) · Science · doi:10.1126/science.aah6524 · Cited by 1,206
Female teachers’ math anxiety affects girls’ math achievement
Beilock, S. L., Gunderson, E. A., Ramirez, G., & Levine, S. C. (2010) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.0910967107 · Cited by 1,116
Deutsch, F. M. (2007) · Gender & Society · doi:10.1177/0891243206293577 · Cited by 870
Gender Differences in Risk Assessment: Why do Women Take Fewer Risks than Men?
Harris, C. R., & Jenkins, M. (2006) · Judgment and Decision Making · doi:10.1017/s1930297500000346 · Cited by 652
Bias in algorithmic filtering and personalization
Bozdag, E. (2013) · Ethics and Information Technology · doi:10.1007/s10676-013-9321-6 · Cited by 649
The Gender Stereotyping of Emotions
Plant, E. A., Hyde, J. S., Keltner, D., & Devine, P. G. (2000) · Psychology of Women Quarterly · doi:10.1111/j.1471-6402.2000.tb01024.x · Cited by 545
Lavy, V. (2008) · Journal of Public Economics · doi:10.1016/j.jpubeco.2008.02.009 · Cited by 290
The Political Biases of ChatGPT
Rozado, D. (2023) · Social Sciences · doi:10.3390/socsci12030148 · Cited by 281 · cited 1×
“ChatGPT has been found to be left-leaning in terms of political viewpoints (e.g., Rodazo 2023).”
BackgroundCave, S., & Dihal, K. (2020) · Philosophy & Technology · doi:10.1007/s13347-020-00415-6 · Cited by 279
Pink or blue: Environmental gender stereotypes in the first two years of life
Pomerleau, A., Bolduc, D., Malcuit, G., & Cossette, L. (1990) · Sex Roles · doi:10.1007/bf00288339 · Cited by 279
The New Politics of Gender Equality
Squires, J. (2007) · doi:10.1007/978-1-137-03653-7 · Cited by 237
Girls' math performance under stereotype threat: The moderating role of mothers' gender stereotypes.
Tomasetto, C., Alparone, F. R., & Cadinu, M. (2011) · Developmental Psychology · doi:10.1037/a0024047 · Cited by 222
Gender Bias in Leader Evaluations: Merging Implicit Theories and Role Congruity Perspectives
Hoyt, C. L., & Burnette, J. L. (2013) · Personality and Social Psychology Bulletin · doi:10.1177/0146167213493643 · Cited by 185
On Artificial Intelligence and Manipulation
Ienca, M. (2023) · Topoi · doi:10.1007/s11245-023-09940-3 · Cited by 113
Responsible AI Systems: Who are the Stakeholders?
Deshpande, A., & Sharp, H. (2022) · Proceedings of the 2022 AAAI/ACM Conference on AI, Ethics, and Society · doi:10.1145/3514094.3534187 · Cited by 103
Banchefsky, S., & Park, B. (2018) · Social Sciences · doi:10.3390/socsci7020027 · Cited by 94
Increasing Gender Diversity in Higher Education Leadership: The Role of Executive Search Firms
Manfredi, S., Clayton-Hathway, K., & Cousens, E. (2019) · Social Sciences · doi:10.3390/socsci8060168 · Cited by 49
Does hype create irreversibilities? Affective circulation and market investments in digital health
Geiger, S., & Gross, N. (2017) · Marketing Theory · doi:10.1177/1470593117692024 · Cited by 42
Self-citationIs ChatGPT Biased? A ReviewPreprint
Singh, S., & Ramakrishnan, N. (2023) · OSF Preprints (OSF Preprints) · doi:10.31219/osf.io/9xkbu · Cited by 35
Current Achievements, Challenges and Digital Chances of Knowledge Based Economy
(2020) · Lecture notes in networks and systems · doi:10.1007/978-3-030-47458-4 · Cited by 23
When Good Algorithms Go Sexist: Why and How to Advance AI Gender Equity
Smith, G., & Rustagi, I. (2021) · Stanford Social Innovation Review · doi:10.48558/a179-b138 · Cited by 19
Patterning of Sexual Violence against Women across US Cities and Counties
Pabst, J., Walfield, S. M., & Schacht, R. N. (2022) · Social Sciences · doi:10.3390/socsci11050208 · Cited by 4
Research works printed but not linked: 13
- Asdal, Kristin, and Reinertsen, Hilde (2021). Doing Document Analysis: A Practice-Oriented Method, Sage Publications.
- Austin, John Langshaw (1962). How to Do Things with Words, Clarendon Press.
- Butler, Judith (1990). Gender Trouble: Feminism and the Subversion of Identity, Routledge.
- Butler, Judith (1993). Bodies That Matter: On the Discursive Limits of “Sex”, Routledge. Routledge Classics.
- Butler, 1994, Gender as Performance: An Interview with Judith Butler, Radical Philosophy: A Journal of Socialist and Feminist Philosophy, 67, 32
- Butler, 2009, Performativity, Precarity and Sexual Politics, AIBR: Revista de Antropología Iberoamericana, 4, i
- Attridge, Derek (1992). Acts of Literature, Routledge.
- Ferrera, Emilio (2023). Should ChatGPT be Biased? Challenges and Risks of Bias in Large Language Models. arXiv.
- Glosh, Sourojit, and Caliskan, Aylin (2023). ChatGPT Perpetuates Gender Bias in Machine Translation and Ignores Non-Gendered Pronouns: Findings across Bengali and Five other Low-Resource Languages. arXiv.
- OECD (2023, June 20). The Pursuit of Gender Equality: An Uphill Battle. Available online: https://www.oecd-ilibrary.org/social-issues-migration-health/the-pursuit-of-gender-equality_9789264281318-en.
- Rothchild, Jennifer (2023, June 21). Gender Bias. In The Blackwell Encyclopedia of Sociology. Available online: https://onlinelibrary.wiley.com/doi/10.1002/9781405165518.wbeosg011.pub2. · doi:10.1002/9781405165518.wbeosg011.pub2
- Shah, Chirag, and Bender, Emily M. (, January March). Situating Search. Paper presented at ACM SIGIR Conference on Human Information Interaction and Retrieval (CHIIR ‘22), Regensburg, Germany.
- Shen, Xinyue, Chen, Zeyuan, Backes, Michael, and Zhang, Yang (2023). In ChatGPT We Trust? Measuring and Characterizing the Reliability of ChatGPT. arXiv.
Cited by
Raman, R., Mandal, S., Das, P., Kaur, T., Sanjanasri, J. P., & Nedungadi, P. (2024) · Human Behavior and Emerging Technologies · doi:10.1155/2024/3085910 · Cited by 214
Kofinas, A. K., Tsay, C. H., & Pike, D. (2025) · British Journal of Educational Technology · doi:10.1111/bjet.13585 · Cited by 207
ChatGPT and mental healthcare: balancing benefits with risks of harms
Blease, C., & Torous, J. (2023) · BMJ Mental Health · doi:10.1136/bmjment-2023-300884 · Cited by 122
Large Language Model‐Based Chatbots in Higher Education
Yigci, D., Eryilmaz, M., Yetisen, A. K., Tasoglu, S., & Ozcan, A. (2025) · Advanced Intelligent Systems · doi:10.1002/aisy.202400429 · Cited by 113
Blease, C., Worthen, A., & Torous, J. (2024) · Psychiatry Research · doi:10.1016/j.psychres.2024.115724 · Cited by 113
“This "algorithmic discrimination" risks perpetuating or worsening inequities in healthcare (Teno, 2023) with evidence that responses exhibit gender, race, and disability biases (Gross, 2023; King, 2022).”
Metadata
- Authors
- Gross, N.
- Affiliations
- National College of Ireland (IE)
- Corresponding author
- Gross, Nicole
- ORCID
- Gross, Nicole
- Venue
- Social Sciences
- Publisher
- MDPI AG
- Published
- 2023-08-01
- Volume
- 12
- Issue
- 8
- Pages
- 435
- ISSN
- 2076-0760 (ISSN-L)
- OpenAlex
- W4385497937
- Cited by
- 162
- References
- 32
- Citations by year
- 2026: 34 · 2025: 62 · 2024: 57 · 2023: 8
- FWCI
- 40.20
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- Locations
- Social Sciences (publishedVersion, cc-by)
- NORMA (submittedVersion)
- RePEc: Research Papers in Economics (submittedVersion)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Ethics and Social Impacts of AI
- Social Sciences › Gender Studies › Gender and Feminist Studies
- Social Sciences › Arts and Humanities › History and Philosophy of Science › Bioethics and Human Rights Issues
- Keywords
- ChatGPT, gender bias, large language models, performativity, gender stereotypes, gender scripts, gendering, undoing gender, gender inequality, generative AI, gender diversity
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.