The Political Biases of ChatGPT
Open accessThe DOI matches its Crossref record; 12 of the 17 references it cites were found in a registry or an index, and 5 are not checked yet.
Abstract
Recent advancements in Large Language Models (LLMs) suggest imminent commercial applications of such AI systems where they will serve as gateways to interact with technology and the accumulated body of human knowledge. The possibility of political biases embedded in these models raises concerns about their potential misusage. In this work, we report the results of administering 15 different political orientation tests (14 in English, 1 in Spanish) to a state-of-the-art Large Language Model, the popular ChatGPT from OpenAI. The results are consistent across tests; 14 of the 15 instruments diagnose ChatGPT answers to their questions as manifesting a preference for left-leaning viewpoints. When asked explicitly about its political preferences, ChatGPT often claims to hold no political opinions and to just strive to provide factual and neutral information. It is desirable that public facing artificial intelligence systems provide accurate and factual information about empirically verifiable issues, but such systems should strive for political neutrality on largely normative questions for which there is no straightforward way to empirically validate a viewpoint. Thus, ethical AI systems should present users with balanced arguments on the issue at hand and avoid claiming neutrality while displaying clear signs of political bias in their content.
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 cited source says what the claim attributes to it.
In the citing paper
“Available data point to ChatGPT output expressing more progressive than traditional political views [38][39][40].”
From The Impact of Artificial Intelligence on Human Sexuality: A Five-Year Literature Review 2020–2024 (2024) · cites this paper as [38]
The cited source says what the claim attributes to it.
In the citing paper
“It shared findings from a study where 15 different political orientation tests were administered to ChatGPT. The results were significant, with 14 out of the 15 tests indicating that ChatGPT's responses inclined towards left-leaning viewpoints.”
From Generative AI in Higher Education: The ChatGPT Effect (2024) · cites this paper as Rozado (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
Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., Kaiser, L., & Polosukhin, I. (2017) · arXiv · doi:10.48550/arxiv.1706.03762 · Cited by 120,404 · cited 1×
“In particular, Large Language Models (LLMs) (Zhou et al. 2022) based on the Transformer architecture (Vaswani et al. 2017) have pushed the state-of-the-art substantially in natural language tasks such as machine translation (Dabre et al. 2020), sentiment analysis (Ain et al. 2017), name entity…”
BackgroundA Survey on Deep Learning for Named Entity Recognition
Li, J., Sun, A., Han, J., & Li, C. (2022) · IEEE Transactions on Knowledge and Data Engineering · doi:10.1109/tkde.2020.2981314 · Cited by 1,567
Chatbots: History, technology, and applications
Adamopoulou, E., & Moussiades, L. (2020) · Machine Learning with Applications · doi:10.1016/j.mlwa.2020.100006 · Cited by 1,090 · cited 1×
“…architecture (Vaswani et al. 2017) have pushed the state-of-the-art substantially in natural language tasks such as machine translation (Dabre et al. 2020), sentiment analysis (Ain et al. 2017), name entity recognition (Li et al. 2022), or dialogue bots (Adamopoulou and Moussiades 2020).”
BackgroundDeep Learning vs. Traditional Computer Vision
O’Mahony, N., Campbell, S., Carvalho, A., Harapanahalli, S., Hernandez, G. V., Krpalkova, L., Riordan, D., & Walsh, J. (2020) · Advances in Intelligent Systems and Computing · doi:10.1007/978-3-030-17795-9_10 · Cited by 1,025
Sentiment Analysis Using Deep Learning Techniques: A Review
Tul, Q., Ali, M., Riaz, A., Noureen, A., Kamranz, M., Hayat, B., & Rehman, A. (2017) · International Journal of Advanced Computer Science and Applications · doi:10.14569/ijacsa.2017.080657 · Cited by 334 · cited 1×
“…architecture (Vaswani et al. 2017) have pushed the state-of-the-art substantially in natural language tasks such as machine translation (Dabre et al. 2020), sentiment analysis (Ain et al. 2017), name entity recognition (Li et al. 2022), or dialogue bots (Adamopoulou and Moussiades 2020).”
BackgroundAlgorithmic Bias: From Discrimination Discovery to Fairness-aware Data Mining
Hajian, S., Bonchi, F., & Castillo, C. (2016) · Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining · doi:10.1145/2939672.2945386 · Cited by 334 · cited 1×
“The topic of algorithmic bias has received an increasing amount of attention in the machine learning academic literature ( Wikipedia 2023a; Kirkpatrick 2016; Cowgill and Tucker 2017; Garcia 2016; Hajian et al. 2016).”
BackgroundLarge Language Models Are Human-Level Prompt Engineers
Zhou, Y., Muresanu, A. I., Han, Z., Paster, K., Pitis, S., Chan, H., & Ba, J. (2022) · arXiv · doi:10.48550/arxiv.2211.01910 · Cited by 330 · cited 1×
“In particular, Large Language Models (LLMs) (Zhou et al. 2022) based on the Transformer architecture (Vaswani et al. 2017) have pushed the state-of-the-art substantially in natural language tasks such as machine translation (Dabre et al. 2020), sentiment analysis (Ain et al. 2017), name entity…”
BackgroundA Survey of Multilingual Neural Machine Translation
Dabre, R., Chu, C., & Kunchukuttan, A. (2021) · ACM Computing Surveys · doi:10.1145/3406095 · Cited by 294 · cited 2×
“The 2012–2022 decade has witnessed spectacular improvements in AI, from computer vision (O’Mahony et al. 2020), to machine translation (Dabre et al. 2020), to generative models for images ( Wikipedia 2023c) and text ( Wikipedia 2022).”
BackgroundGarcia, M. (2016) · World Policy Journal · doi:10.1215/07402775-3813015 · Cited by 257 · cited 1×
“The topic of algorithmic bias has received an increasing amount of attention in the machine learning academic literature ( Wikipedia 2023a; Kirkpatrick 2016; Cowgill and Tucker 2017; Garcia 2016; Hajian et al. 2016).”
BackgroundThe American Journalist in the Digital Age: Another Look at U.S. News People
Weaver, D. H., Willnat, L., & Wilhoit, G. C. (2019) · Journalism & Mass Communication Quarterly · doi:10.1177/1077699018778242 · Cited by 174 · cited 1×
“…of professionals working in these institutions are politically left-leaning (Reuters Institute for the Study of Journalism n.d.; Hopmann et al. 2010; Weaver et al. 2019; Langbert 2018; Archive et al. 2021; Schoffstall 2022; American Enterprise Institute—AEI (blog) n.d.; The Harvard Crimson n.d.).”
Background
Research works printed but not linked: 2
- Bender, Emily M., Gebru, Timnit, McMillan-Major, Angelina, and Shmitchell, Shmargaret (2021). FAccT ’21: Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency, Association for Computing Machinery.
- Hajian, Sara, Bonchi, Francesco, and Castillo, Carlos (2016). KDD ’16: Proceedings of the 22Nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, ACM.
Cited by
Grassini, S. (2023) · Education Sciences · doi:10.3390/educsci13070692 · Cited by 1,048
“It has also been found that, in some cases, the information delivered by ChatGPT may be biased politically [57,58]; as well on religion, race, gender, and fairness (for a review see [59]).”
BackgroundChatGPT and Open-AI Models: A Preliminary Review
Roumeliotis, K. I., & Tselikas, N. D. (2023) · Future Internet · doi:10.3390/fi15060192 · Cited by 706
“Research/Classification/Academic Writing Social/Politics ChatGPT’s responses are left-leaning political viewpoints [66].”
BackgroundSynthetic Replacements for Human Survey Data? The Perils of Large Language Models
Bisbee, J., Clinton, J. D., Dorff, C., Kenkel, B., & Larson, J. M. (2024) · Political Analysis · doi:10.1017/pan.2024.5 · Cited by 209
“…of public opinion and the creation of synthetic data, connecting with a growing body of research that asks similar questions (Bender et al. 2021; Spirling 2023) and documents similar issues (Abdulhai et al. 2023; Cao et al. 2023; Motoki, Pinho Neto, and Rodrigues 2023; Rozado 2023).”
Generative AI in Higher Education: The ChatGPT Effect
Chan, C. K. Y., & Colloton, T. (2024) · Routledge · doi:10.4324/9781003459026 · Cited by 203
The Self-Perception and Political Biases of ChatGPT
Rutinowski, J., Franke, S., Endendyk, J., Dormuth, I., & Pauly, M. (2023) · arXiv · doi:10.48550/arxiv.2304.07333 · Cited by 178
“Compared to the publications that conducted similar tests with ChatGPT [24], [25], [28], [29], we also obtain results indicating a political bias of ChatGPT towards progressive views.”
ResultKofinas, A. K., Tsay, C. H., & Pike, D. (2025) · British Journal of Educational Technology · doi:10.1111/bjet.13585 · Cited by 207
What ChatGPT Tells Us about Gender: A Cautionary Tale about Performativity and Gender Biases in AI
Gross, N. (2023) · Social Sciences · doi:10.3390/socsci12080435 · Cited by 162
“ChatGPT has been found to be left-leaning in terms of political viewpoints (e.g., Rodazo 2023).”
BackgroundChatGPT: Challenges and Benefits in Software Programming for Higher Education
Silva, C. A. G. D., Ramos, F. N., de Moraes, R. V., & Santos, E. L. D. (2024) · Sustainability · doi:10.3390/su16031245 · Cited by 149
“This can have a negative impact on research studies and educational materials that rely on ChatGPT-generated content [45–47].”
BackgroundAkpan, I. J., Kobara, Y. M., Owolabi, J., Akpan, A. A., & Offodile, O. F. (2025) · International Transactions in Operational Research · doi:10.1111/itor.13522 · Cited by 137
The limitations and ethical considerations of ChatGPT
Hua, S., Jin, S., & Jiang, S. (2024) · Data Intelligence · doi:10.1162/dint_a_00243 · Cited by 120
“Books1 and Books2 used by GPT-3 are actually not detailed in the original work [61], but some related researchers speculated [66], that Books1 used by GPT-3 is the Gutenberg data from SPGC (Standardized Project Gutenberg Corpus) and Books2 may be a dataset composed from the Bibliotik e-book website…”
BackgroundThe Social Impact of Generative AI: An Analysis on ChatGPT
Baldassarre, M. T., Caivano, D., Fernandez Nieto, B., Gigante, D., & Ragone, A. (2023) · Proceedings of the 2023 ACM Conference on Information Technology for Social Good · doi:10.1145/3582515.3609555 · Cited by 112
Wan, T., & Chen, Z. (2024) · Physical Review Physics Education Research · doi:10.1103/physrevphyseducres.20.010152 · Cited by 109
“For example, AI-generated feedback could be potentially biased [34,35]because LLMs like GPT are pre-trained from large amount of data from different sources including the internet and social media, and therefore its outputs can be influenced by existing bias contain in the source of its training.”
BackgroundA Systematic Review of the Limitations and Associated Opportunities of ChatGPT
Cong-Lem, N., Soyoof, A., & Tsering, D. (2024) · International Journal of Human–Computer Interaction · doi:10.1080/10447318.2024.2344142 · Cited by 101
The Self-Perception and Political Biases of ChatGPT
Rutinowski, J., Franke, S., Endendyk, J., Dormuth, I., Roidl, M., & Pauly, M. (2024) · Human Behavior and Emerging Technologies · doi:10.1155/2024/7115633 · Cited by 98
Pellert, M., Lechner, C. M., Wagner, C., Rammstedt, B., & Strohmaier, M. (2024) · Perspectives on Psychological Science · doi:10.1177/17456916231214460 · Cited by 97
“Researchers have noticed examples of apparent “sani-tization” or other corrections of model text outputs happening behind the scenes to models available only through APIs or web interfaces (Rozado, 2023).”
The Security of Using Large Language Models: A Survey with Emphasis on ChatGPT
Zhou, W., Zhu, X., Han, Q.-L., Li, L., Chen, X., Wen, S., & Xiang, Y. (2025) · IEEE/CAA Journal of Automatica Sinica · doi:10.1109/jas.2024.124983 · Cited by 95
Hayes, A. S. (2025) · International Journal of Qualitative Methods · doi:10.1177/16094069251322346 · Cited by 184
“To be sure, LLM outputs re fl ect learned data patterns rather than deliberate malice (see: Guo et al., 2024; Rozado, 2023).”
The political preferences of LLMs
Rozado, D. (2024) · PLOS ONE · doi:10.1371/journal.pone.0306621 · Cited by 83
“The left-leaning political preferences of ChatGPT have been documented previously [11].”
BackgroundMethodSelf-citationDiminished diversity-of-thought in a standard large language model
Park, P. S., Schoenegger, P., & Zhu, C. (2024) · Behavior Research Methods · doi:10.3758/s13428-023-02307-x · Cited by 83
“…consistent with the finding that the strength and direction of ChatGPT’s political leanings – as measured by certain survey questions – varied over time (Rozado, 2023), as OpenAI used public feedback on the chatbot to make closed-source updates between the release dates of the chatbot’s multiple…”
BackgroundLi, L., Peng, W., & Rheu, M. M. J. (2024) · Telemedicine and e-Health · doi:10.1089/tmj.2023.0313 · Cited by 79
“67 These factors could all complicate individuals’ attitudes toward chatbots.”
Bukar, U. A., Sayeed, M. S., Fatimah Abdul Razak, S., Yogarayan, S., & Sneesl, R. (2024) · IEEE Access · doi:10.1109/access.2024.3425172 · Cited by 69
“The digital divide [21], [29], [31], [37], [98], [99] can be exacerbated by unequal access to AI technologies, leading to disparities in educational opportunities.”
BackgroundStudents' reflections on their experience with ChatGPT
Šedlbauer, J., Činčera, J., Slavík, M., & Hartlová, A. (2024) · Journal of Computer Assisted Learning · doi:10.1111/jcal.12967 · Cited by 71
Introduction to large language models (LLMs) for dementia care and research
Treder, M. S., Lee, S., & Tsvetanov, K. A. (2024) · Frontiers in Dementia · doi:10.3389/frdem.2024.1385303 · Cited by 63
“Rozado (2023) administered multiple political orientation tests to ChatGPT.”
Polyportis, A., & Pahos, N. (2024) · Humanities and Social Sciences Communications · doi:10.1057/s41599-023-02464-6 · Cited by 64
Bodroza, B., Dinic, B. M., & Bojic, L. (2023) · arXiv · doi:10.48550/arxiv.2306.04308 · Cited by 59
The Impact of Artificial Intelligence on Human Sexuality: A Five-Year Literature Review 2020–2024
Döring, N., Le, T. D., Vowels, L. M., Vowels, M. J., & Marcantonio, T. L. (2024) · Current Sexual Health Reports · doi:10.1007/s11930-024-00397-y · Cited by 65
“While some researchers are concerned about AI re-affirming traditional worldviews including, for example, conservative gender roles (e.g., [37]), others are concerned about AI being implemented in such a way that it supports progressive, left-leaning worldviews (e.g., [38]).”
Zhu, S., Wang, Z., Zhuang, Y., Jiang, Y., Guo, M., Zhang, X., & Gao, Z. (2024) · Telematics and Informatics Reports · doi:10.1016/j.teler.2024.100138 · Cited by 66
Shafik, W. (2023) · Advances in computational intelligence and robotics book series · doi:10.4018/979-8-3693-0502-7.ch001 · Cited by 56
Farina, M., Yu, X., & Lavazza, A. (2024) · AI and Ethics · doi:10.1007/s43681-023-00405-2 · Cited by 74
“…written about the limitations of self-super-vised training, and it is well-known that chatbots (such as ChatGPT) may suffer from biases due to the political and value orientations of those who have financed their development or own the infrastructure required for their functioning [20, 27, 39, 51].”
Systems engineering issues for industry applications of large language model
Chen, W., Yan-yi, L., Tie-zheng, G., Li, D., Tao, H., Li, Z., Yang, Q., Hui-han, W., & Ying-you, W. (2023) · Applied Soft Computing · doi:10.1016/j.asoc.2023.111165 · Cited by 78
Kim, J., Lee, J., Jang, K. M., & Lourentzou, I. (2023) · Telematics and Informatics · doi:10.1016/j.tele.2023.102085 · Cited by 63
Yang, Z., Wu, J. G., & Xie, H. (2024) · Asia Pacific Journal of Education · doi:10.1080/02188791.2023.2300137 · Cited by 75
Alvero, A. J., Lee, J., Regla-Vargas, A., Kizilcec, R. F., Joachims, T., & Antonio, A. L. (2024) · Journal of Big Data · doi:10.1186/s40537-024-00986-7 · Cited by 48
“However, the self-evident potential of LLMs has raised critical questions about their biases, tendency to emulate certain political and moral stances, and ability to fabricate references when used as a research assistant [1, 73].”
BackgroundSpennemann, D. H. R. (2024) · Heritage · doi:10.3390/heritage7030070 · Cited by 58
A Ghostwriter for the Masses: ChatGPT and the Future of Writing
Zimmerman, A. (2023) · Annals of Surgical Oncology · doi:10.1245/s10434-023-13436-0 · Cited by 44
Strengths, Weaknesses, Opportunities, and Threats of Using ChatGPT in Scientific Research
Giray, L., Jacob, J., & Gumalin, D. L. (2024) · International Journal of Technology in Education · doi:10.46328/ijte.618 · Cited by 44
An integrative decision-making framework to guide policies on regulating ChatGPT usage
Bukar, U. A., Sayeed, M. S., Abdul Razak, S. F., Yogarayan, S., & Ahmed Amodu, O. (2024) · PeerJ Computer Science · doi:10.7717/peerj-cs.1845 · Cited by 43
“…of the studies encompass exploratory approaches ( Haluza & Jungwirth, 2023 ; Yeadon et al., 2023 ; Victor et al., 2023 ; Dwivedi et al., 2023 ; Rozado, 2023 ; Cascella et al., 2023 ; Ariyaratne et al., 2023 ; Yan, 2023 ; Grünebaum et al., 2023 ; Kooli, 2023 ; Jungwirth & Haluza, 2023 ;…”
BackgroundInfluentialAli, H., & Aysan, A. F. (2024) · International Journal of Ethics and Systems · doi:10.1108/ijoes-04-2024-0112 · Cited by 64
Islam, M. B. E., Haseeb, M., Batool, H., Ahtasham, N., & Muhammad, Z. (2024) · Blockchains · doi:10.3390/blockchains2040020 · Cited by 77
Revisiting the political biases of ChatGPT
Fujimoto, S., & Takemoto, K. (2023) · Frontiers in Artificial Intelligence · doi:10.3389/frai.2023.1232003 · Cited by 42
“This limitation can introduce bias into the test results ( Rozado, 2023a).”
BackgroundMethodResultInfluentialSalinas, A., Shah, P., Huang, Y., McCormack, R., & Morstatter, F. (2023) · ACM Conference on Equity and Access in Algorithms, Mechanisms, and Optimization (EAAMO) · doi:10.1145/3617694.3623257 · Cited by 118
Role and challenges of ChatGPT and similar generative artificial intelligence in arts and humanities
Rane, N., & Choudhary, S. (2024) · Studies in Humanities and Education · doi:10.48185/she.v5i1.999 · Cited by 59
Assessing trustworthy AI: Technical and legal perspectives of fairness in AI
Kattnig, M., Angerschmid, A., Reichel, T., & Kern, R. (2024) · Computer law & security review · doi:10.1016/j.clsr.2024.106053 · Cited by 77
Generative AI in Marketing: Promises, Perils, and Public Policy Implications
Kumar, V., Kotler, P., Gupta, S., & Rajan, B. (2024) · Journal of Public Policy & Marketing · doi:10.1177/07439156241286499 · Cited by 96
“…gender, race, religion, and other social constructs (e.g., Wan et al. 2023); (3) users ’ writing and thinking being in fl uenced through opinionated outputs (Jakesch et al. 2023); and (4) left-leaning bias (in politics) that can impact policy making (Motoki, Neto, and Rodrigues 2024; Rozado 2023).”
Generative AI Meets Open-Ended Survey Responses: Research Participant Use of AI and Homogenization
Zhang, S., Xu, J., & Alvero, A. (2025) · Sociological Methods & Research · doi:10.1177/00491241251327130 · Cited by 101
Cantens, T. (2024) · AI & Society · doi:10.1007/s00146-023-01840-9 · Cited by 77
“Similar questions are emerging on political neutrality, as recent research explores the “political biases” of GenAI (Motoki et al. 2023; Rozado 2023), calling for “ethical” GenAI to maintain its political “neutrality”.”
Synthetic Replacements for Human Survey Data? The Perils of Large Language Models
Bisbee, J., Clinton, J. D., Dorff, C., Kenkel, B., & Larson, J. (2023) · doi:10.31235/osf.io/5ecfa · Cited by 213
ChatGPT-4 as a journalist: Whose perspectives is it reproducing?
Breazu, P., & Katsos, N. (2024) · Discourse & Society · doi:10.1177/09579265241251479 · Cited by 51
“Biases in Large Language Models (LLMs) can be conceptualised as unintentional echoes of the societal worldviews present in their training data (Gross, 2023; Rozado, 2023).”
Bodroža, B., Dinić, B. M., & Bojić, L. (2024) · Royal Society Open Science · doi:10.1098/rsos.240180 · Cited by 95
ChatGPT in Academic Writing: A Scientometric Analysis of Literature Published Between 2022 and 2023
Lendvai, G. F. (2025) · Journal of Empirical Research on Human Research Ethics · doi:10.1177/15562646251350203 · Cited by 57
“…purposes, as the platform regularly exhibits content and opinions that are diagnostic, educational, cultural, demographic, political and ideo-logical, and in some instances, even gender bias (Gross, 2023; Huang, 2024; Motoki et al., 2023; Qu & Wang, 2024; Rozado, 2023; Schmidt et al., 2024).”
Xu, Z., Fang, Q., Huang, Y., & Xie, M. (2024) · PLoS ONE · doi:10.1371/journal.pone.0302502 · Cited by 52
“Word cloud visualization intuitively presents these high-frequency terms, making critical information easily accessible [43].”
BackgroundPrandner, D., Wetzelhütter, D., & Hese, S. (2025) · Frontiers in Education · doi:10.3389/feduc.2024.1417900 · Cited by 56
Misinformation, Disinformation, and Generative AI: Implications for Perception and Policy
Jaidka, K., Chen, T., Chesterman, S. A., Hsu, W., Kan, M., Kankanhalli, M., Lee, M. L., Seres, G., Sim, T., Taeihagh, A., et al. (2024) · Digital Government Research and Practice · doi:10.1145/3689372 · Cited by 75
ChatGPT: towards AI subjectivity
D’Amato, K. (2024) · AI & Society · doi:10.1007/s00146-024-01898-z · Cited by 41
“Directly stemming from this structural fixity, the pre-vailing values in the text corpora, and the conversation-oriented training, ChatGPT is open to a range of criticisms, the commonest being an attribution of bias; e.g. that it is manifestly “left-leaning” (Rozado 2023) or “woke” (McGee 2023).”
Chen, Z., & Wan, T. (2025) · Physical Review Physics Education Research · doi:10.1103/physrevphyseducres.21.010126 · Cited by 57
Llama 2: Early Adopters' Utilization of Meta's New Open-Source Pretrained ModelPreprint
Roumeliotis, K. I., Tselikas, N. D., & Nasiopoulos, D. Κ. (2023) · Preprints.org · doi:10.20944/preprints202307.2142.v2 · Cited by 51
AI’s Epistemic Harm: Reinforcement Learning, Collective Bias, and the New AI Culture War
Humphreys, D. (2025) · Philosophy & Technology · doi:10.1007/s13347-025-00928-y · Cited by 26
“This will follow from research which has shown that language models are not politically neutral (Bang et al., 2024 ; Feng et al., 2023; Rozado, 2023, 2024 ).”
Emerging AI individualism: how young people integrate social AI into everyday life
Brandtzaeg, P. B., Følstad, A., & Skjuve, M. (2025) · Communication and Change · doi:10.1007/s44382-025-00011-2 · Cited by 30
ChatGPT giving advice on how to cheat in university assignments: how workable are its suggestions?
Spennemann, D. H. R., Biles, J. M., Brown, L., Ireland, M. F., Longmore, L., Singh, C., Wallis, A., & Ward, C. (2024) · Interactive Technology and Smart Education · doi:10.1108/itse-10-2023-0195 · Cited by 37
Can AI Mitigate Bias in Writing Letters of Recommendation?
Leung, T. I., Sagar, A., Shroff, S., & Henry, T. L. (2023) · JMIR Medical Education · doi:10.2196/51494 · Cited by 28
“LLMs have already been shown to, for example, generate statements that have certain political leanings [32,33] or have cultural biases [34,35].”
BackgroundEffect of Gratitude Expression of AI Chatbot on Willingness to Donate
Namkoong, M., Park, G., Park, Y., & Lee, S. (2023) · International Journal of Human-Computer Interaction · doi:10.1080/10447318.2023.2259719 · Cited by 34
“For example, in the case of ChatGPT, it is known that left-leaning answers are preferred to users’ questions (Rozado, 2023).”
Spennemann, D. H. R. (2025) · Publications · doi:10.3390/publications13010012 · Cited by 23
“…political orientation tests showed that ChatGPT is biassed and ChatGPT exhibits a preference for libertarian, progressive, and left-leaning viewpoints (Hartmann et al., 2023; McGee, 2023; Motoki et al., 2023; Rozado, 2023; Rutinowski et al., 2023), with a North American slant (Cao et al., 2023).”
ChatGPT: A Threat Against the CIA Triad of Cyber Security
Chowdhury, M. M., Rifat, N., Ahsan, M., Latif, S., Gomes, R., & Rahman, M. S. (2023) · IEEE International Conference on Electro Information Technology · doi:10.1109/eit57321.2023.10187355 · Cited by 39
“But on the other hand, researchers [15] discuss how the increasing use of Large Language Models (LLMs) as gateways to technology and human knowledge raises concerns about political biases in these models.”
BackgroundGhafouri, V., Agarwal, V., Zhang, Y., Sastry, N. R., Such, J., & Suárez‐Tangil, G. (2023) · ACM International Conference on Information and Knowledge Management (CIKM) · doi:10.1145/3583780.3614777 · Cited by 481
“Rozado [19] have applied 15 political orientation tests to ChatGPT by prompting us-ing the test’s style to engineer the exact prompt for ChatGPT (see Table 1 for a sample).”
BackgroundMethodInfluentialCollu, M. G., Salviati, U., Confalonieri, R., Conti, M., & Apruzzese, G. (2025) · arXiv · doi:10.48550/arxiv.2508.20863 · Cited by 22
Llama 2: Early Adopters' Utilization of Meta's New Open-Source Pretrained ModelPreprint
Roumeliotis, K. I., Tselikas, N. D., & Nasiopoulos, D. Κ. (2023) · Preprints.org · doi:10.20944/preprints202307.2142.v1 · Cited by 40
The ethical security of large language models: A systematic review
Liu, F., Jiang, J., Lu, Y., Huang, Z., & Jiang, J. (2025) · Frontiers of Engineering Management · doi:10.1007/s42524-025-4082-6 · Cited by 29
“This is particularly concerning in authoritarian countries, where AI development may be exploited for manipulation (Jungherr, 2023; Motoki et al., 2024; Rozado, 2023; Saetra, 2023; Sang and Yu, 2023; Zhang, 2023).”
Natural language processing for social science research: A comprehensive review
Hou, Y., & Huang, J. (2025) · Chinese Journal of Sociology · doi:10.1177/2057150x241306780 · Cited by 45
“Rozado (2023) conducted a multilingual political bias test on ChatGPT and found that it tends to hold left-leaning views, which may be related to the fact that the training materials for large language models are mainly sourced from the Internet, which is dominated by in fl uential institutions in…”
Auditing GPT's Content Moderation Guardrails: Can ChatGPT Write Your Favorite TV Show?
Mahomed, Y., Crawford, C. M., Gautam, S., Friedler, S. A., & Metaxa, D. (2024) · ACM Conference on Fairness, Accountability, and Transparency (FAccT) · doi:10.1145/3630106.3658932 · Cited by 38
Mapping the individual, social and biospheric impacts of Foundation Models
Hernández, A. D., Krishna, S., Perini, A., Katell, M. A., Bennett, S., Borda, A., Hashem, Y., Hadjiloizou, S., Mahomed, S., Jayadeva, S., et al. (2024) · ACM Conference on Fairness, Accountability, and Transparency (FAccT) · doi:10.1145/3630106.3658939 · Cited by 47
Stability Analysis of ChatGPT-Based Sentiment Analysis in AI Quality Assurance
Ouyang, T., MaungMaung, A., Konishi, K., Seo, Y., & Echizen, I. (2024) · Electronics · doi:10.3390/electronics13245043 · Cited by 26
Does Artificial Intelligence Cause More Harm than Good in Schools?
Zainuddin, N. (2024) · International Journal of Language Education and Applied Linguistics · doi:10.15282/ijleal.v14i1.10432 · Cited by 26
“These biases can show up in their educational uses (Rozado, 2023).”
BackgroundWe are Building Gods: AI as the Anthropomorphised Authority of the Past
Öhman, C. (2024) · Minds and Machines · doi:10.1007/s11023-024-09667-z · Cited by 34
Generative Artificial Intelligence and the Future of Public Knowledge
Spennemann, D. H. R. (2025) · Knowledge · doi:10.3390/knowledge5030020 · Cited by 18
“ChatGPT, for example, often purports to merely strive to provide factual and neutral information and not to hold political opinions [156].”
Synthetic users: insights from designers’ interactions with persona-based chatbots
Gu, H., Chandrasegaran, S. K., & Lloyd, P. A. (2025) · Artificial intelligence for engineering design analysis and manufacturing · doi:10.1017/s0890060424000283 · Cited by 34
“Notably, studies have identified significant political (Motoki et al., 2024; Rozado, 2023) and social biases (Zack et al., 2024; Howard and Borenstein, 2018) in models accessible to the public, potentially reflecting and amplifying existing user prejudices in a manner similar to existing media…”
Goh, T., Dai, X., & Yang, Y. (2023) · BenchCouncil Transactions on Benchmarks Standards and Evaluations · doi:10.1016/j.tbench.2024.100153 · Cited by 20
Philosophical Investigations into AI Alignment: A Wittgensteinian Framework
Pérez‐Escobar, J. A., & Sarikaya, D. (2024) · Philosophy & Technology · doi:10.1007/s13347-024-00761-9 · Cited by 23
“For instance, ChatGPT commits to specific political positions in a way qualified as algorithmic bias (e.g., McGee, 2023; Motoki et al., 2023; Rozado, 2023; Rutinowski et al., 2023), but this may be the intentional outcome of human political bias or the result of misalignment.”
Maltby, J., Rayes, T., Nage, A., Sharif, S., Omar, M., & Nichani, S. (2024) · PLoS ONE · doi:10.1371/journal.pone.0307164 · Cited by 41
“However, despite improvements [44], it still retains some level of political bias influenced by factors like the languages used in the system [45–47].”
BackgroundDöring, N. (2023) · Zeitschrift für Sexualforschung · doi:10.1055/a-2142-9527 · Cited by 14
Slow Writing with ChatGPT: Turning the Hype into a Right Way Forward
Sirisathitkul, C. (2023) · Postdigital Science and Education · doi:10.1007/s42438-023-00441-5 · Cited by 19
“…references in ChatGPT’s output have been well-documented (Ariyaratne et al. 2023; Day 2023; Kim 2023; Metze et al. 2023; Walters and Wilde 2023), and reports have also indicated instances of bias expressed by ChatGPT (Chatterjee and Dethlefs 2023; Farrokhnia et al. 2023; Ray 2023; Rozado 2023).”
“How Does ChatGPT Work” Examining Functionality To The Creative AI CHATGPT on X's (Twitter) Platform
Alahmed, Y., Abadla, R., Badri, A. A., & Ameen, N. (2023) · International Conference on Social Networks Analysis, Management and Security (SNAMS) · doi:10.1109/snams60348.2023.10375450 · Cited by 17
“Some early chatbots, like Eliza, managed to pass the Turing test using clever tricks, though they didn't exhibit substantial intelligence [2].”
BackgroundMethodDoes artificial intelligence bias perceptions of environmental challenges?
van der Ven, H., Corry, D., Elnur, R. I., Provost, V. J., Syukron, M., & Tappauf, N. (2024) · Environmental Research Letters · doi:10.1088/1748-9326/ad95a2 · Cited by 24
Laflamme, A. S., & Bruneault, F. (2025) · Journal of Scholarly Publishing · doi:10.3138/jsp-2024-1125 · Cited by 48
“One must also acknowledge that different language models may exhibit distinct political stances and therefore are likely to generate politically oriented content (Rozado, 2023 ).”
Goodbye human annotators? Content analysis of social policy debates using ChatGPT
Gielens, E., Sowula, J., & Leifeld, P. (2025) · Journal of Social Policy · doi:10.1017/s0047279424000382 · Cited by 20
“Data privacy concerns as well as possible climate impacts (e.g. water usage; see Li et al., 2024) and potential political bias (McGee, 2023; Rozado, 2023) must not be neglected when using ChatGPT.”
Large Language Models and Financial Market SentimentPreprint
Bond, S. A., Klok, H., & Zhu, M. (2023) · SSRN Electronic Journal · doi:10.2139/ssrn.4584928 · Cited by 15
Doing Thematic Analysis in the Age of Generative AI: Practices, Ethics and Reflexivity
Xu, W. J. (2026) · International Journal of Qualitative Methods · doi:10.1177/16094069261425173 · Cited by 20
Hayes, A. S. (2023) · doi:10.31235/osf.io/yms8p · Cited by 12
Utilizing Text-Generative AI for Creating Oral Reading Fluency Probes
Sidwell, M. D., Bonner, L. W., Bates‐Brantley, K., & Wu, S. (2024) · Intervention in School and Clinic · doi:10.1177/10534512241235896 · Cited by 11
“Furthermore, despite rigorous training of the model, language learning models can produce outputs influenced by the biases of its developers, unintended or otherwise (Ray, 2023; Rozado, 2023).”
ChatGPT: the challenges it presents for writing assignments
Meishar-Tal, H. (2024) · TechTrends · doi:10.1007/s11528-024-00972-z · Cited by 10
“This phenomenon can be attributed to ChatGPT’s tendency to generate texts reflecting biases toward English-speaking and liberal Western cultural perspectives (McGee, 2023; Ray, 2023; Rozado, 2023).”
ChatGPT in Higher Education: Perceptions of Computer Science-Related Students
Hammer, S., Ottinger, S., Zönnchen, B., Hohendanner, M., Hobelsberger, M., & Thurner, V. (2024) · IEEE Global Engineering Education Conference (EDUCON) · doi:10.1109/educon60312.2024.10578742 · Cited by 15
“Since ChatGPT was diagnosed to be politically biased [17], [18] and its opinionated responses can affect what users write and think [19], utilizing it prompts a challenge for students and educators alike.”
BackgroundGeorgiou, G. P. (2024) · Journal of Language and Social Psychology · doi:10.1177/0261927x241298337 · Cited by 16
“LLMs demonstrate various biases associated with gender (Gross, 2023), language (Georgiou, 2024), religion (Abid et al., 2021), politics (Rozado, 2023), and nationality (Venkit et al., 2023) among others.”
Kamińska, A. M. (2026) · The Journal of Academic Librarianship · doi:10.1016/j.acalib.2026.103218 · Cited by 12
AI chatbots and political learning
Taylor, J. B., & Richey, S. (2024) · Journal of Information Technology & Politics · doi:10.1080/19331681.2024.2422929 · Cited by 10
“Using ChatGPT responses to fourteen political orientation tests in English and one in Spanish, evidence suggests its responses are generally on the political left (Rozado, 2023).”
“Turning right”? An experimental study on the political value shift in large language models
Liu, Y., Panwang, Y., & Gu, C. (2025) · Humanities and Social Sciences Communications · doi:10.1057/s41599-025-04465-z · Cited by 31
“The test questionnaire, consisting of 62 questions, is one of the most commonly used political compass tools, as demonstrated in previous studies (Motoki et al. 2023; Fujimoto and Takemoto 2023; Rozado 2023; Hartmann et al. 2023; Falck et al. 2020; Lan and Li, 2015).”
Demirkol, M., & Malkoç, N. (2023) · Educational Process International Journal · doi:10.22521/edupij.2023.124.3 · Cited by 11
“The analysis showed that the central theme of Large-Language-Models (0.85, 0.5) had relationships with the Student-Character, Political-Bias, Health, Gpt-3.”
ResultInfluentialExploring Perspectives of Teacher Students Toward Generative AI Technologies
Thararattanasuwan, K., & Prachagool, V. (2024) · International Education Studies · doi:10.5539/ies.v17n5p22 · Cited by 14
“The notion that artificial intelligence systems such as ChatGPT may provide findings that are biased and unjust was there (Hassani & Silva, 2023; Rozado, 2023).”
BackgroundPrompting the Machine: Introducing an LLM Data Extraction Method for Social Scientists
Foisy, L.-O. M., Proulx, É., Cadieux, H., Gilbert, J., Rivest, J., Bouillon, A., & Dufresne, Y. (2025) · Social Science Computer Review · doi:10.1177/08944393251344865 · Cited by 13
“This concern grows with the rapid expansion of accessible AI systems and our increasing reliance on them (Rozado, 2023).”
Noels, S., Bied, G., Buyl, M., Rogiers, A., Fettach, Y., Lijffijt, J., & Bie, T. D. (2025) · Lecture notes in computer science · doi:10.1007/978-3-032-05962-8_16 · Cited by 36
Thekdi, S., & Aven, T. (2024) · Journal of Risk Research · doi:10.1080/13669877.2024.2340013 · Cited by 7
“For example, Rozado (2023) conducts political orientation tests on the information system and concludes ‘left-leaning’ political orientation.”
Comparing ChatGPT with human judgements of social traits from face photographs
Kramer, R. S. S. (2025) · Computers in Human Behavior Artificial Humans · doi:10.1016/j.chbah.2025.100156 · Cited by 11
Metadata
- Authors
- Rozado, D.
- Affiliations
- Te Pūkenga (NZ)
- Corresponding author
- Rozado, David
- ORCID
- Rozado, David
- Venue
- Social Sciences
- Publisher
- MDPI AG
- Published
- 2023-03-02
- Volume
- 12
- Issue
- 3
- Pages
- 148
- ISSN
- 2076-0760 (ISSN-L)
- OpenAlex
- W4323043839
- Cited by
- 281
- References
- 12
- Citations by year
- 2026: 63 · 2025: 81 · 2024: 98 · 2023: 39
- FWCI
- 70.11
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- Locations
- Social Sciences (publishedVersion, cc-by)
- 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 › Sociology and Political Science › Misinformation and Its Impacts
- Social Sciences › Law › Artificial Intelligence in Law
- Keywords
- ChatGPT, political bias, large language models, political neutrality, AI ethics, normative questions
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.