Bias of AI-generated content: an examination of news produced by large language models
Open accessPublished version of a preprint
The DOI matches its Crossref record; 36 of the 36 references it cites were found in a registry or an index.
Abstract
Large language models (LLMs) have the potential to transform our lives and work through the content they generate, known as AI-Generated Content (AIGC). To harness this transformation, we need to understand the limitations of LLMs. Here, we investigate the bias of AIGC produced by seven representative LLMs, including ChatGPT and LLaMA. We collect news articles from The New York Times and Reuters, both known for their dedication to provide unbiased news. We then apply each examined LLM to generate news content with headlines of these news articles as prompts, and evaluate the gender and racial biases of the AIGC produced by the LLM by comparing the AIGC and the original news articles. We further analyze the gender bias of each LLM under biased prompts by adding gender-biased messages to prompts constructed from these news headlines. Our study reveals that the AIGC produced by each examined LLM demonstrates substantial gender and racial biases. Moreover, the AIGC generated by each LLM exhibits notable discrimination against females and individuals of the Black race. Among the LLMs, the AIGC generated by ChatGPT demonstrates the lowest level of bias, and ChatGPT is the sole model capable of declining content generation when provided with biased prompts.
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
Figures










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
“In LLM research, bias direction varies by model and task, with some studies finding pro-Black and others anti-Black patterns depending on the context [3, 29, 46].”
From All Public Voices Are Equal, But Are Some More Equal Than Others to LLMs? (2026) · cites this paper as [29]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Findings on direction are mixed: An et al. [3] found consistent pro-female bias, while others find anti-female patterns depending on task [29].”
From All Public Voices Are Equal, But Are Some More Equal Than Others to LLMs? (2026) · cites this paper as [29]
Experimental — accuracy not yet measuredThe cited source says what the claim attributes to it.
In the citing paper
“Neither race nor gender showed consistent effects, a result that diverges from the existing literature's primary focus on these two dimensions [3, 29, 94].”
From All Public Voices Are Equal, But Are Some More Equal Than Others to LLMs? (2026) · cites this paper as [29]
The cited source says what the claim attributes to it.
In the citing paper
“Additionally, LLM-generated content often exhibits inherent biases [20] and can be misused for malicious purposes [37,53], further complicating efforts to maintain information integrity.”
From Beyond Binary: Towards Fine-Grained LLM-Generated Text Detection via Role Recognition and Involvement Measurement (2025) · cites this paper as [20]
In the citing paper
“AI algorithms that have been trained on biased data may disseminate false information or reinforce prejudices (Fang et al., 2024).”
From AI literacy in media education: A comparative study of mass communication students in the UAE and Pakistan (2025) · cites this paper as Fang et al., 2024
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
(2003) · Journal of Machine Learning Research · doi:10.5555/944919.944937 · Cited by 32,337
Dissecting racial bias in an algorithm used to manage the health of populations
Obermeyer, Z., Powers, B., Vogeli, C., & Mullainathan, S. (2019) · Science · doi:10.1126/science.aax2342 · Cited by 7,137
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
An Introduction to Categorical Data Analysis
Agresti, A. (2007) · Wiley Series in Probability and Statistics · doi:10.1002/0470114754 · Cited by 6,689
LLaMA: Open and Efficient Foundation Language Models
Touvron, H., Lavril, T., Izacard, G., Martinet, X., Lachaux, M.-A., Lacroix, T., Rozière, B., Goyal, N., Hambro, E., Azhar, F., et al. (2023) · arXiv · doi:10.48550/arxiv.2302.13971 · Cited by 4,602
The Earth Mover's Distance as a Metric for Image Retrieval
Rubner, Y., Tomasi, C., & Guibas, L. J. (2000) · International Journal of Computer Vision · doi:10.1023/a:1026543900054 · Cited by 4,568
Training language models to follow instructions with human feedback
Ouyang, L., Wu, J., Jiang, X., Almeida, D., Wainwright, C. L., Mishkin, P., Zhang, C., Agarwal, S., Slama, K., Ray, A., et al. (2022) · arXiv · doi:10.48550/arxiv.2203.02155 · Cited by 4,208
Pennycook, G., & Rand, D. G. (2021) · Trends in Cognitive Sciences · doi:10.1016/j.tics.2021.02.007 · Cited by 1,360 · cited 1×
“In future research, considering that people tend to trust news based on the news source and names of the journalist 22 , it would be intriguing to explore whether an LLM generates more or less biased content when the prompt includes both the news title and journalist name.”
BackgroundFriedman, B., & Nissenbaum, H. (1996) · ACM Transactions on Information Systems · doi:10.1145/230538.230561 · Cited by 1,203 · cited 1×
“At the sentence level, the bias of AIGC is evaluated as the difference between sentences related to each investigated population group in AIGC and their counterparts in the news articles produced by the selected news agencies, in terms of their expressed sentiments and toxicities.”
BackgroundStereoSet: Measuring stereotypical bias in pretrained language models
Nadeem, M., Bethke, A., & Reddy, S. (2021) · Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers) · doi:10.18653/v1/2021.acl-long.416 · Cited by 643
Research works printed but not linked: 5
- Baker RS Hawn A Algorithmic bias in education Int. J. Artif. Intell. Educ. 2021 32 1 41
- Liang, P. P., Wu, C., Morency, L. -P. & Salakhutdinov, R. Towards understanding and mitigating social biases in language models, in International Conference on Machine Learning, 6565–6576 (PMLR, 2021).
- Nadeem, A., Abedin, B. & Marjanovic, O. Gender bias in AI: A review of contributing factors and mitigating strategies. ACIS 2020 Proc. (2020).
- Řehůřek, R., & Sojka, P. Software framework for topic modelling with large corpora, in Proceedings of the LREC 2010 Workshop on New Challenges for NLP Frameworks, 45–50 (ELRA, Valletta, Malta, 2010).
- Sharpe D Chi-square test is statistically significant: Now what? Pract. Assessment Res. Eval. 2019 20 8
Cited by
A survey on large language model (LLM) security and privacy: The Good, The Bad, and The Ugly
Yao, Y., Duan, J., Xu, K., Cai, Y., Sun, Z., & Zhang, Y. (2024) · High-Confidence Computing · doi:10.1016/j.hcc.2024.100211 · Cited by 1,083
“In this section, we examine the range of existing defense methods against various attacks and vulnerabilities associated with LLMs 1 .”
BackgroundGe, J., Sun, S., Owens, J., Galvez, V., Gologorskaya, O., Lai, J. C., Pletcher, M. J., & Lai, K. (2024) · Hepatology · doi:10.1097/hep.0000000000000834 · Cited by 178
Swisher, A. R., Wu, A. W., Liu, G. C., Lee, M., Carle, T. R., & Tang, D. M. (2024) · Otolaryngology · doi:10.1002/ohn.927 · Cited by 91
Mei, P., Brewis, D. N., Nwaiwu, F., Sumanathilaka, D., Alva-Manchego, F., & Demaree-Cotton, J. (2025) · Computers in Human Behavior: Artificial Humans · doi:10.1016/j.chbah.2025.100140 · Cited by 50
“More widely it has been noted that there is a bias aGenAInst AI use, which is often manifested in hesitancy among individuals to fully accept AI-generated intellectual, artistic and creative works due to the perception of it being "superhuman" (Fang et al., 2024; Ragot et al., 2020).”
Jang, M. (2024) · Informatics · doi:10.3390/informatics11030054 · Cited by 45
Challenges of Implementing LLMs in Clinical Practice: Perspectives
Artsi, Y., Sorin, V., Glicksberg, B. S., Korfiatis, P. D., Freeman, R., Nadkarni, G. N., & Klang, E. (2025) · Journal of Clinical Medicine · doi:10.3390/jcm14176169 · Cited by 66
“Bias in LLM outputs driven by uneven training data and unbalanced prompt design poses a risk, particularly for marginalized populations [27,28].”
Beyond principlism: practical strategies for ethical AI use in research practices
Lin, Z. (2025) · AI and Ethics · doi:10.1007/s43681-024-00585-5 · Cited by 40
“For example, LLMs may generate text that reinforces stereotypes about gender, race [ 37 ], and geographic regions (e.g., associating them with negative attributes like low intelligence or next generation of scientists by not engaging with AI tools.”
How was my performance? Exploring the role of anchoring bias in AI-assisted decision making
Carter, L. D., & Liu, D. (2025) · International Journal of Information Management · doi:10.1016/j.ijinfomgt.2025.102875 · Cited by 64
Pantano, E., Serravalle, F., & Priporas, C. (2024) · Journal of Marketing Management · doi:10.1080/0267257x.2024.2436096 · Cited by 57
“These align with findings from previous studies on the usage of GAI and LLMs to generate new solutions (Fang et al., 2024; Matz et al., 2024; Sarstedt et al., 2024), which can be biased or not completely reflective of human needs or desires.”
'I'm Categorizing LLM as a Productivity Tool': Examining Ethics of LLM Use in HCI Research Practices
Kapania, S., Wang, R., Li, T. J.-J., Li, T., & Shen, H. (2025) · Proceedings of the ACM on Human-Computer Interaction · doi:10.1145/3711000 · Cited by 81
AI-Generated News Content: The Impact of AI Writer Identity and Perceived AI Human-Likeness
Lee, D. C., Jhang, J., & Baek, T. H. (2025) · International Journal of Human-Computer Interaction · doi:10.1080/10447318.2025.2477739 · Cited by 43
“Large language models (LLMs) can generate coherent and contextually relevant text based on input prompts, enabling the creation of news articles, summaries, and even interactive conversations (Fang et al., 2024).”
An outline of Prognostics and health management Large Model: Concepts, Paradigms, and challenges
Tao, L., Li, S., Liu, H., Huang, Q., Ma, L., Ning, G., Chen, Y., Wu, Y., Li, B., Zhang, W., et al. (2025) · Mechanical Systems and Signal Processing · doi:10.1016/j.ymssp.2025.112683 · Cited by 44
On the verge of a digital divide in the use of generative AI?
Suárez, D., & García-Mariñoso, B. (2025) · Telecommunications Policy · doi:10.1016/j.telpol.2025.102997 · Cited by 46
Strzelecki, A. (2024) · Learned Publishing · doi:10.1002/leap.1650 · Cited by 39
“Therefore, the creators of this tool have pointed out its susceptibility to inaccuracies and its occasional tendency to provide mis-leading information (Fang et al., 2024; Polyportis & Pahos, 2024).”
Generative artificial intelligence: implications for biomedical and health professions education
Hersh, W. (2025) · Annual Review of Biomedical Data Science · doi:10.1146/annurev-biodatasci-103123-094756 · Cited by 28
Co-designing Large Language Model Tools for Project-Based Learning with K12 Educators
Ravi, P., Masla, J., Kakoti, G., Lin, G. C., Anderson, E., Taylor, M., Ostrowski, A. K., Breazeal, C., Klopfer, E. D., & Abelson, H. (2025) · CHI Conference on Human Factors in Computing Systems · doi:10.1145/3706598.3713971 · Cited by 59
Biased AI writing assistants shift users’ attitudes on societal issues
Williams‐Ceci, S., Jakesch, M., Bhat, A., Kadoma, K., Zalmanson, L., & Naaman, M. (2026) · Science Advances · doi:10.1126/sciadv.adw5578 · Cited by 61
“…writing may steer the stance of AI suggestions to an extent ( 44 ), biases can also be deliberately introduced into AI suggestions in adversarial scenarios with people providing biased training data, prompts, or feedback during the training process when creating these applications ( 42 , 45 – 47 ).”
Programming with AI: Evaluating ChatGPT, Gemini, AlphaCode, and GitHub Copilot for Programmers
Siam, K., Gu, H., & Cheng, J. Q. (2024) · International Conference on Computing Advancements (ICCA) · doi:10.1145/3723178.3723224 · Cited by 39
Bai, X., & Yang, L. F. (2025) · Frontiers in Psychology · doi:10.3389/fpsyg.2025.1644209 · Cited by 34
Shaping future sustainable cities with AI-powered urban informatics: Toward human-AI symbiosis
Yue, Y., Yan, G., Lan, T., Cao, R., Gao, Q., Gao, W., Huang, B., Huang, G., Huang, Z., Kan, Z., et al. (2025) · Computational Urban Science · doi:10.1007/s43762-025-00190-0 · Cited by 27
“The gap under-scores the need for domain-specific AI innovations and a shift toward unified frameworks inspired by the success of large language models (LLMs) to model urban complexities effectively.”
Zhao, Y., Tang, S., Zhang, H., & Lyu, L. (2025) · Journal of Retailing and Consumer Services · doi:10.1016/j.jretconser.2025.104400 · Cited by 39
Self-citationAllan, K., Azcona, J., Sripada, S. G., Leontidis, G., Sutherland, C. A. M., Phillips, L. H., & Martin, D. S. (2025) · Royal Society Open Science · doi:10.1098/rsos.241472 · Cited by 28
Fan, X., Xiao, Q., Zhou, X., Pei, J., Sap, M., Lu, Z., & Shen, H. (2025) · CHI Conference on Human Factors in Computing Systems · doi:10.1145/3706598.3713477 · Cited by 66
Lotto, C., Sheppard, S. C., Anschuetz, W., Stricker, D., Molinari, G., Huwendiek, S., & Anschuetz, L. (2024) · OTO Open · doi:10.1002/oto2.70018 · Cited by 23
“16 This can lead to outdated examination material without regard to the latest research or guidelines.”
BackgroundInternational AI Safety Report
Bengio, Y., Mindermann, S., Privitera, D., Besiroglu, T., Bommasani, R., Casper, S., Choi, Y., Fox, P., Garfinkel, B., Goldfarb, D., et al. (2025) · arXiv · doi:10.48550/arxiv.2501.17805 · Cited by 16
Haider, S. A., Prabha, S., Gomez-Cabello, C. A., Borna, S., Genovese, A., Trabilsy, M., Elegbede, A. I., Yang, J. F., Galvao, A., Tao, C., et al. (2025) · Journal of Clinical Medicine · doi:10.3390/jcm14103484 · Cited by 15
AI and Epistemic Risk for Democracy: A Coming Crisis of Public Knowledge?Preprint
Wihbey, J. (2024) · SSRN Electronic Journal · doi:10.2139/ssrn.4805026 · Cited by 27
Advancements in AI-Generated Content Forensics: A Systematic Literature Review
Xu, Q., Mu, W., Li, J., Sun, T., & Jiang, X. (2025) · ACM Computing Surveys · doi:10.1145/3760526 · Cited by 26
Using Artificial Intelligence for Scholarly Writing
Oermann, M. H., Owens, J. K., Carter‐Templeton, H., Peterson, G. M., & Bailey, H. E. (2025) · AJN American Journal of Nursing · doi:10.1097/ajn.0000000000000179 · Cited by 23
“24 They found that the content included significant gender and racial bias against women and Black people.”
Political biases and inconsistencies in bilingual GPT models—the cases of the U.S. and China
Zhou, D., & Zhang, Y. (2024) · Scientific Reports · doi:10.1038/s41598-024-76395-w · Cited by 42
Benchmarking the Hallucination Tendency of Google Gemini and Moonshot KimiPreprint
Shan, R., Ming, Q., Hong, G., & Wu, H. (2024) · OSF Preprints (OSF Preprints) · doi:10.31219/osf.io/83rq9 · Cited by 50
Leon, M., Ruaengsri, C., Pelletier, G., Bethencourt, D., Shibata, M., Flores, M. Q., & Shudo, Y. (2024) · Journal of Clinical Medicine · doi:10.3390/jcm13216543 · Cited by 12
“A study evaluating gender and racial biases in content from seven large language models, including ChatGPT, identified substantial biases [58].”
BackgroundThe threat of AI chatbot responses to crowdsourced open-ended survey questions
Traylor, F. (2024) · Energy Research & Social Science · doi:10.1016/j.erss.2024.103857 · Cited by 21
Salih, S. H., Husain, O., Almohamedh, R. M., tajelsier, H., Hashim, A. H. A., Elshafie, H., & Motwakel, A. (2025) · Array · doi:10.1016/j.array.2025.100630 · Cited by 29
Xiao, Q., Fan, X., Simon, F. M., Zhang, B., & Eslami, M. (2025) · CHI Conference on Human Factors in Computing Systems · doi:10.1145/3706598.3714090 · Cited by 35
Ethical Issues in AI-Generated Texts: A Systematic Review and Analysis
Zhang, Y., Zhou, T., Qiao, H., & Li, T. X. (2025) · International Journal of Human-Computer Interaction · doi:10.1080/10447318.2025.2530071 · Cited by 24
“The research conducted by Fang et al. (2024) examined and compared the gender and racial bias in news articles generated by LLMs and reported in the New York Times and Reuters.”
Li, F., Ya, Y., & Yu, G. (2025) · Emerging Media · doi:10.1177/27523543251317572 · Cited by 25
“While generative AI holds remarkable potential, it is far from perfect and often exhibits biases, such as discrimination against women and Black communities — issues that even advanced models like ChatGPT have not fully resolved (Fang et al., 2024).”
He, K., Liu, X., Xu, Y., Bustamante, A. S., & Warschauer, M. (2025) · ACM International Conference on Interaction Design and Children (IDC) · doi:10.1145/3713043.3727056 · Cited by 16
An explainable framework for assisting the detection of AI-generated textual content
Yan, S., Wang, Z., & Dobolyi, D. G. (2025) · Decision Support Systems · doi:10.1016/j.dss.2025.114498 · Cited by 17
On LLM Wizards: Identifying Large Language Models' Behaviors for Wizard of Oz Experiments
Fang, J., Aréchiga, N., Namikoshi, K., Bravo, N. S., Hogan, C., & Shamma, D. A. (2024) · ACM International Conference on Intelligent Virtual Agents (IVA) · doi:10.1145/3652988.3673967 · Cited by 18
Garín, D., Cook, S., Ferry, C., Bennar, W., Togni, M., Meier, P., Wenaweser, P., Puricel, S., & Arroyo, D. (2025) · European Heart Journal - Digital Health · doi:10.1093/ehjdh/ztaf068 · Cited by 11
“…partly because the importance of prompt engineering (i.e. 20 While it is well documented that these models can exhibit medical, gender, or racial biases when generating recommendations for simulated data, 21–23 their potential biases in decision-making for real-world patients remain understudied.”
The Energy Hunger of AI: Large Language Models as Challenges and Enablers for Sustainable Energy
Hinov, N. L. (2025) · Energies · doi:10.3390/en18174701 · Cited by 19
Fairframe: a fairness framework for bias detection and mitigation in news
Sallami, D., & Aïmeur, E. (2024) · AI and Ethics · doi:10.1007/s43681-024-00568-6 · Cited by 19
“In this research, bias refers to the phenomenon where computer systems “systematically and unfairly discriminate against certain individuals or groups of individuals in favor of others" [83].”
The Role of Large Language Model Chatbots in Sexual Education: An Unmet Need of Research
Mondal, H., & Mondal, S. (2025) · Journal of Psychosexual Health · doi:10.1177/26318318251323714 · Cited by 13
A systematic review of sentiment analytics in banking headlines
Jayanthakumaran, M., Shukla, N., Pradhan, B., & Beydoun, G. (2025) · Decision Analytics Journal · doi:10.1016/j.dajour.2025.100584 · Cited by 12
A method for generating personalized learning content based on AIGC
Hu, Z. (2025) · Sustainable Futures · doi:10.1016/j.sftr.2025.101331 · Cited by 8
DOSA: A Dataset of Social Artifacts from Different Indian Geographical Subcultures
Seth, A., Ahuja, S., Bali, K., & Sitaram, S. (2024) · doi:10.2139/ssrn.4756716 · Cited by 8
Toal, M. P., Hill, C. J., Quinn, M., O’Neill, C., & Maxwell, A. P. (2025) · Journal of Medical Internet Research · doi:10.2196/73603 · Cited by 10
“LLMs can generate false information in the form of hallucinations and produce gender-or racially biased outputs [2,12,13].”
Carbajal-Carrera, B. (2024) · Australian Review of Applied Linguistics · doi:10.1075/aral.24077.car · Cited by 13
Ali, S., Hassan, D., Hegazy, A., Selim, N., & Khadragy, S. (2025) · Online Journal of Communication and Media Technologies · doi:10.30935/ojcmt/17661 · Cited by 7
Navigating Generative AI Disclosure, Ownership, and Accountability in Co-Creative Domains
Do, H. J., Feldman, M. Q., He, J., Hwang, A. H., & Kim, S. (2025) · Annual Symposium on Human-Computer Interaction for Work (CHIWORK) · doi:10.1145/3707640.3729212 · Cited by 8
Perception vs reality: a multi-level framework for analyzing user perceptions of LLM-based chatbots
Liu, Y., & Rau, P. P. (2025) · Information Technology and People · doi:10.1108/itp-03-2024-0403 · Cited by 9
How are faculty and college students embracing AI? — A multi-informant mixed method study
Xie, L., Jiang, Y., Chang, C., Zeng, X.-Y., Hong, J. S., & Mo, F. (2025) · Computers and Education Artificial Intelligence · doi:10.1016/j.caeai.2025.100506 · Cited by 10
Bias in GenAI: Exploring cultural variations between large language models
Zhang, Y., Tan, L. L. W., & O’Halloran, K. L. (2026) · Linguistics and Education · doi:10.1016/j.linged.2026.101515 · Cited by 16
Shan, K., Patel, M. A., McCreary, M., Punnen, T. G., Villalobos, F., Tardo, L. M., Horton, L. A., Sguigna, P. V., Blackburn, K. M., Munoz, S. B., et al. (2024) · Journal of the Neurological Sciences · doi:10.1016/j.jns.2024.123360 · Cited by 7
Beyond principlism: Practical strategies for ethical AI use in research practicesPreprint
Lin, Z. (2023) · SSRN Electronic Journal · doi:10.2139/ssrn.4576581 · Cited by 5
Zhang, H., Maillo, A., Khan, S. A., Martinez‐de‐Morentin, X., Lehmann, R., Gómez-Cabrero, D., & Tegner, J. N. (2024) · Computational and Structural Biotechnology Journal · doi:10.1016/j.csbj.2024.10.034 · Cited by 7
“Meanwhile, incorporating insuf-ficiently validated LLMs into the peer review process could result in biased outputs [134,135].”
Self-citationKhapre, S., Mersha, M. A., Shakil, H., Baruah, J., & Kalita, J. K. (2025) · Expert Systems with Applications · doi:10.1016/j.eswa.2025.129832 · Cited by 15
“For instance, AI-generated content can reflect and amplify racial, gender, and political biases present in the training corpus, perpetuating toxic discourse [150].”
Defining and assessing AI literacy for researchers across the research lifecycle
Parker, J. L., & Becker, K. P. (2026) · Frontiers in Education · doi:10.3389/feduc.2026.1827603 · Cited by 5
Molina, J. C. C. (2024) · Derecom Derecho de la Comunicación · doi:10.5209/dere.98112 · Cited by 12
Does ChatGPT show gender bias in behavior detection?
Wu, J., Song, Y., & Wu, D. C. (2024) · Humanities and Social Sciences Communications · doi:10.1057/s41599-024-04219-3 · Cited by 7
“For example, some studies using arti fi cially constructed test cases found that ChatGPT falls short in terms of gender equality and exhibits consistency issues across different versions (Geiger et al., 2024; Fang et al., 2024).”
Li, S., Lin, X., Liu, Y., Chen, X., & Li, J. (2025) · IEEE Communications Magazine · doi:10.1109/mcom.004.2400646 · Cited by 17
“This can amplify social biases, especially in sensitive areas like news generation, where underrepresented groups may face discrimination [7].”
Artificial Intelligence and the Interpretation of the Past
Magnani, M., & Clindaniel, J. (2025) · Advances in Archaeological Practice · doi:10.1017/aap.2025.10110 · Cited by 6
“Racial, gender-based, and regional biases threaten being baked into emergent programs (Fang et al. 2024; Roselli et al. 2019).”
How Reliable AI Chatbots are for Disease Prediction from Patient Complaints?
Nipu, A. S., Islam, K. M. S., & Madiraju, P. (2024) · IEEE International Conference on Information Reuse and Integration (IRI) · doi:10.1109/iri62200.2024.00052 · Cited by 10
“We select Gemini Ultra 1.0 as part of our disease prediction task.”
MethodInfluentialAnthony, A., Sharma, S., & Sharma, S. (2025) · Bulletin of Science Technology & Society · doi:10.1177/02704676251353105 · Cited by 12
“Ethical issues such as plagiarism, inherent bias, and inadequate citation practices demand critical attention (Fang et al., 2024; Ray, 2023).”
AIGC in Design: Critical Thinking Challenges and Opportunities Revealed Through Systematic Review
Ge, P., Meng, D., & Fan, F. (2025) · Lecture notes in computer science · doi:10.1007/978-3-031-93429-2_2 · Cited by 8
Cheng, Z., Zhou, L., Jiang, F., Wang, B., & Li, H. (2025) · Proceedings of the ACM on Web Conference 2025 · doi:10.1145/3696410.3714770 · Cited by 3
Giunti, G., Glynn, R. W., Hennessy, J., & Doherty, C. P. (2025) · Journal of Medical Internet Research · doi:10.2196/73173 · Cited by 3
Conversational agents and equity in health care systems: a scoping review
Morias, A., & Renmans, D. (2026) · Journal of Health Equity · doi:10.1080/29944694.2025.2606723 · Cited by 3
Ruggiano, N., Sahoo, S., Brashear, A., Nwatu, U. L., Brunson, A., Noh, H., Cole, H. S., McKinney, R., Suarez, C. V. F., Brown, E. L., et al. (2026) · JMIR Medical Education · doi:10.2196/83085 · Cited by 5
“Research has found that content generated through AI platforms can, at times, be falsely fabricated (ie, hallucinations) or can draw on content that is outdated or biased [37].”
Bhatt, K. N., AlAshry, M. S., Kohli, S., Kabia, S. K., & Seabra, C. (2026) · International Journal of Information Management · doi:10.1016/j.ijinfomgt.2026.103058 · Cited by 4
Academic Psychiatry in the Age of Artificial Intelligence
King, D. R., Liu, H. Y., & Brenner, A. M. (2025) · Academic Psychiatry · doi:10.1007/s40596-025-02112-y · Cited by 4
“Finally, most foundational models upon which GAI tools are built have been shown to be gender and racially biased, with most instances of harm occurring against Black women in particular [25, 26].”
Iqhrammullah, M., Gusti, N., Muzaffar, A., Khader, Y. S., Maulana, S., Rademaker, M., & Abdullah, A. (2025) · Health Policy and Technology · doi:10.1016/j.hlpt.2025.101058 · Cited by 7
Seth, I., Lim, B., Phan, R., Xie, Y., Kenney, P. S., Bukret, W. E., Thomsen, J. B., Cuomo, R., Ross, R. J., Ng, S., et al. (2024) · Medicina · doi:10.3390/medicina60091500 · Cited by 5
“Secondly, algorithms must be refined to minimise biases and enhance the models’ ability to provide accurate and reliable medical recommendations [21–25].”
BackgroundWhy AI image generators cannot afford to be blind to racial bias
Arif, M., & Takefuji, Y. (2025) · AI & Society · doi:10.1007/s00146-025-02258-1 · Cited by 4
Impact of YouTube User‐Generated Content on News Dissemination and Youth Information Reception
Chunqiong, W., Shan, J., Jianhong, S., & Yingqi, L. (2025) · Health Expectations · doi:10.1111/hex.70408 · Cited by 6
“UGC reinforces gender roles and vocational aspirations among youth, but these studies lacked cross ‐ cultural generalisability and overlooked news ‐ literacy implications [13, 19, 27].”
Sui, M. (2025) · AI & Society · doi:10.1007/s00146-025-02628-9 · Cited by 7
“3 See more details here: https:// github. com/ dalab udel/ llmbi as. 4 Fang et al. (2024) investigated the text generated by multiple LLMs (i.e., GPT-2, GPT-3, GPT-4, and Cohere), but the analysis in this pre-sent study only employed their GPT-4 generated news stories.”
Ethnic and Gender Representation in German, Italian and Dutch School Textbooks
Jehle, A. M. C., van Veen, D., Groeneveld, M. G., van der Pol, L. D., & Mesman, J. (2025) · Journal of Community Psychology · doi:10.1002/jcop.70063 · Cited by 3
“Similarly, AI ‐ generated content can be biased in how it portrays women (Fang et al. 2024) and ethnically minoritized groups (Fang et al. 2024; Yang 2025).”
Kim, D., Kim, S., Lee, Y., & Kim, C. (2025) · International Journal of Human-Computer Interaction · doi:10.1080/10447318.2025.2598674 · Cited by 2
“In particular, AI technologies may internalize not only gender biases (Hitron et al., 2022) but also racial biases (Fang et al., 2024) owing to their reliance on learning from human-provided data and algorithms.”
Reasoning About Human Values in GitHub Issues: What Can a Large Language Model Reveal?
Mougouei, D., Haque, A., Nurwidyantoro, A., Mougouei, E., Fahmideh, M., Shahin, M., Dharmawan, A., Widodo, T. W., & Dam, H. K. (2026) · International Journal of Human-Computer Interaction · doi:10.1080/10447318.2025.2594137 · Cited by 3
“…in issue discussions, enhancing our practical understanding of human values (Mougouei et al., 2018), it does not eliminate the risk of misinterpreting values, hallucination (confabulation) (Mittal et al., 2024; Song et al., 2025), and the inherent bias in large language models (Fang et al., 2024).”
Hsieh, C., Wang, D., He, Y., & Hsieh, D. W. (2026) · Review of Public Personnel Administration · doi:10.1177/0734371x261417717 · Cited by 2
Zhang, J., Cai, Y., Xiang, Y., & Sun, D. (2024) · Lecture Notes in Education Psychology and Public Media · doi:10.54254/2753-7048/50/20240923 · Cited by 2
“[ 4]AIGC can extract information from vast data sources, automatically create news content that meets grammatical and semantic standards, and optimize the text to some extent, achieving automation, diversification, and customization in news content generation.”
InfluentialAssessing Gender and Racial Bias in Large Language Model‐Powered Virtual Reference
Liu, J., & Wang, H. (2024) · Proceedings of the Association for Information Science and Technology · doi:10.1002/pra2.1061 · Cited by 6
“Although LLMs currently have many potential biases, researchers are dedicated to reducing these biases and continuously improving the quality of LLMs. Researchers have proposed approaches such as Reinforcement Learning from Human Feedback (RLHF) (Ouyang et al., 2022) to incorporate human judgment and ethical norms int…”
InfluentialOrganizational Readiness for Artificial Intelligence (AI) in Network Security
Greene, B. A., & Burton, S. L. (2024) · Advances in human resources management and organizational development book series · doi:10.4018/979-8-3693-8562-3.ch007 · Cited by 4
Martino, A. S., Miller, M. N., Moumos, E., Trung, R., & White, K. (2025) · The Canadian Journal of Human Sexuality · doi:10.3138/cjhs-2024-0018 · Cited by 4
Deng, J. M., & Wang, S. (2025) · Journal of Global Trends in Social Science · doi:10.70731/4yfd8137 · Cited by 2
GenFair: Systematic Test Generation for Fairness Fault Detection in Large Language Models
Srinivasan, M., & Abdel, J. (2025) · IEEE International Conference on Artificial Intelligence Testing (AITest) · doi:10.1109/aitest66680.2025.00030 · Cited by 2
“However, despite their advances, LLMs often inherit biases from the vast datasets they are trained on, which can result in discriminatory or harmful outcomes, particularly in sensitive areas such as finance and healthcare [3].”
Olivos, F., Kamelski, T., & Gac, S. A. (2025) · Teaching Sociology · doi:10.1177/0092055x251397371 · Cited by 4
“…representation of age, gender, race, and socioeconomic groups (Abid, Farooqi, and Zou 2021; Bai et al. 2025; Bianchi et al. 2023; Chu et al. 2022; Fang et al. 2024; Heikkilä 2023; Hu et al. 2024; Jenka 2023; Kamelski and Klinge 2024; Liu, Panwang, and Gu 2025; Luccioni et al. 2023; Naik and…”
Ding, Z., Wei, R., Xia, J., Mu, Y., Wang, J., & Lin, Y. (2024) · Briefings in Bioinformatics · doi:10.1093/bib/bbae641 · Cited by 3
“Biases in LLMs trained on general text corpora may affect their applicability to specific domains like bioinformatics and biomedical science [122–124], leading to suboptimal results in Ribo-seq data analysis.”
Does Present AI Make Library Services Better?
Verma, V. K. (2025) · Journal of Information and Knowledge · doi:10.17821/srels/2025/v62i1/171702 · Cited by 1
The Ethical Dilemma and Governance Path of Generative AI in Cultural Communication
Mei, B., & Zhu, H. (2024) · doi:10.1145/3722237.3722403 · Cited by 2
Venturing ChatGPT's lens to explore human values in software artifacts: a case study of mobile APIs
Mougouei, D., Rafi, S., Fahmideh, M., Mougouei, E., Khan, J. A., Dam, H. K., Nurwidyantoro, A., & Chaudron, M. R. V. (2025) · Behaviour and Information Technology · doi:10.1080/0144929x.2025.2478278 · Cited by 2
“Studies have shown that LLM-generated content demonstrates gender and racial biases (Fang et al. 2024).”
Chen, K., Yao, R., Tang, Z., Yao, A., & Jin, X. (2025) · Design and Artificial Intelligence · doi:10.1016/j.daai.2025.100004 · Cited by 4
Paltaratskaya, V., Ji, A., Mazumdar, P., & Wise, K. R. (2025) · Journal of Current Issues & Research in Advertising · doi:10.1080/10641734.2025.2498996 · Cited by 2
“On sensitive topics like abortion (Maci and Abbiati 2023), public health (Spitale et al. 2023), and biased news headlines (Fang et al. 2024), LLMs can convincingly generate misinformation.”
Design Method of Aging Service Robot Based on AIGC Incorporating TRIZ Theory
Yang, Y., Wu, L., & Zhao, R. (2025) · Lecture notes in computer science · doi:10.1007/978-3-031-93508-4_25 · Cited by 1
Kılıçkaya, F. (2025) · Advances in computational intelligence and robotics book series · doi:10.4018/979-8-3373-5092-9.ch009 · Cited by 1
Bias in Language Models: Interplay of Architecture and Data?
Talebpour, M., Long, Y., de Herrera, A. G. S., & Jameel, S. (2025) · International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR) · doi:10.1145/3726302.3730172 · Cited by 2
Scripts and Stories: AI's Role in Modern Creative Writing
Syamsuri, A. S., & Asrifan, A. (2025) · Advances in computational intelligence and robotics book series · doi:10.4018/979-8-3373-1057-2.ch006 · Cited by 1
Kahani: Culturally-Nuanced Visual Storytelling Tool for Non-Western Cultures
Hamna, Sudharsan, D., Seth, A., Budhiraja, R., Khullar, D., Jain, V., Bali, K., Vashistha, A., & Segal, S. (2025) · ACM SIGCAS/SIGCHI Conference on Computing and Sustainable Societies (COMPASS) · doi:10.1145/3715335.3735478 · Cited by 10
Can LLMs Beat Humans on Discerning Human-written and LLM-generated Science News?
Soós, D., Jiang, M., & Wu, J. F. (2025) · ACM Conference on Hypertext and Social Media (HT) · doi:10.1145/3720553.3746674 · Cited by 1
Metadata
- Authors
- Fang, X., Che, S., Mao, M., Zhang, H., Zhao, M., & Zhao, X.
- Affiliations
- University of Delaware (US)
- Tsinghua University (CN)
- Chinese University of Hong Kong, Shenzhen (CN)
- Shanghai University of Finance and Economics (CN)
- Corresponding author
- Fang, Xiao
- Che, Shangkun
- Venue
- Scientific Reports
- Published
- 2024-03-04
- Volume
- 14
- Issue
- 1
- Pages
- 5224-5224
- ISSN
- 2045-2322 (ISSN-L)
- PMC
- PMC10909834
- OpenAlex
- W4392362609
- Cited by
- 194
- References
- 28
- Citations by year
- 2026: 65 · 2025: 93 · 2024: 34
- FWCI
- 79.58
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- cc-by
- creativecommons.org/licenses/by/4.0 (tdm)
- creativecommons.org/licenses/by/4.0 (vor)
- Locations
- Scientific Reports (publishedVersion, cc-by)
- Research Square (acceptedVersion, cc-by)
- PubMed Central (publishedVersion, cc-by)
- PubMed (publishedVersion)
- Library, Museums and Press - UDSpace (University of Delaware) (submittedVersion, cc-by)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Ethics and Social Impacts of AI
- Physical Sciences › Computer Science › Artificial Intelligence › Topic Modeling
- Physical Sciences › Computer Science › Artificial Intelligence › Hate Speech and Cyberbullying Detection
- Keywords
- AI-generated content, large language models, algorithmic bias, gender bias, racial bias, ChatGPT, llamas, discrimination against women, Black people, New York Times, Reuters
- MeSH terms
- Animals, Aortic Valve Insufficiency, Camelids, New World, Female, Humans, Language, Male, Bias, Sexism
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.