Persistent Anti-Muslim Bias in Large Language Models
Open accessThe DOI matches its Crossref record; 13 of the 13 references it cites were found in a registry or an index.
Abstract
It has been observed that large-scale language models capture undesirable societal biases, e.g. relating to race and gender; yet religious bias has been relatively unexplored. We demonstrate that GPT-3, a state-of-the-art contextual language model, captures persistent Muslim-violence bias. We probe GPT-3 in various ways, including prompt completion, analogical reasoning, and story generation, to understand this anti-Muslim bias, demonstrating that it appears consistently and creatively in different uses of the model and that it is severe even compared to biases about other religious groups. For instance, Muslim is analogized to terrorist in 23% of test cases, while Jewish is mapped to its most common stereotype, money, in 5% of test cases. We quantify the positive distraction needed to overcome this bias with adversarial text prompts, and find that use of the most positive 6 adjectives reduces violent completions for Muslims from 66% to 20%, but which is still higher than for other religious groups.
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
“Pre-trained on massive amounts of text on the Internet, LMs are known to produce linguistically fawed, factually incorrect, or even ethically problematic text at times [1,24].”
From CoAuthor: Designing a Human-AI Collaborative Writing Dataset for Exploring Language Model Capabilities (2022) · cites this paper as [1]
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
“LLMs and other AI tools may have a number of biases that could arise from various factors, such as unbalanced training datasets74–77.”
From On opportunities and challenges of large multimodal foundation models in education (2025) · cites this paper as [75]
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
“it is also a liability; all AI systems might inherit the same problematic biases of a few foundation models [Bolukbasi et al. 2016; Caliskan et al. 2017; Abid et al. 2021, inter alia]”
From On the Opportunities and Risks of Foundation Models (2021) · cites this paper as Abid et al. 2021
Experimental — accuracy not yet measuredThe cited source says what the claim attributes to it.
In the citing paper
“Previous work has shown that language models tend to exhibit human-like biases with respect to specific protected groups such as gender, ethnicity or religious orientation3–6.”
From Generative language models exhibit social identity biases (2024) · cites this paper as [4]
In the citing paper
“If a medical LLM has been trained on data that contains misinformation or biased content, LLMs deployed for clinical decision-making are at risk of reproducing these problematic contents [120,124–126].”
From Bias in medical AI: Implications for clinical decision-making (2024) · cites this paper as [126]
1 more claim about this paper
Create a free account to see every claim other papers make about this one.
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×
“These languagemodels have evolved, from learning individual word vectors with single-layer models [8], to more complex language generation architectures such as recurrent neural networks [5] and most recently transformers [4, 6, 14].”
BackgroundEfficient Estimation of Word Representations in Vector Space
Mikolov, T., Chen, K., Corrado, G., & Dean, J. (2013) · arXiv · doi:10.48550/arxiv.1301.3781 · Cited by 26,568 · cited 1×
“These languagemodels have evolved, from learning individual word vectors with single-layer models [8], to more complex language generation architectures such as recurrent neural networks [5] and most recently transformers [4, 6, 14].”
BackgroundLanguage Models are Few-Shot Learners
Brown, T. B., Mann, B., Ryder, N., Subbiah, M., Kaplan, J., Dhariwal, P., Neelakantan, A., Shyam, P., Sastry, G., Askell, A., et al. (2020) · arXiv · doi:10.48550/arxiv.2005.14165 · Cited by 23,372 · cited 4×
“These languagemodels have evolved, from learning individual word vectors with single-layer models [8], to more complex language generation architectures such as recurrent neural networks [5] and most recently transformers [4, 6, 14].”
BackgroundMethodInfluentialSelf-Critical Sequence Training for Image Captioning
Rennie, S. J., Marcheret, E., Mroueh, Y., Ross, J., & Goel, V. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.131 · Cited by 2,112 · cited 1×
“These descriptions are generated by passing images through a pretrained image captioning model [11], and fine-tuning GPT-3 to transform the short captions into long descriptions, as shown in Fig.”
MethodReformer: The Efficient Transformer
Kitaev, N., Kaiser, Ł., & Levskaya, A. (2020) · arXiv · doi:10.48550/arxiv.2001.04451 · Cited by 1,435 · cited 1×
“These languagemodels have evolved, from learning individual word vectors with single-layer models [8], to more complex language generation architectures such as recurrent neural networks [5] and most recently transformers [4, 6, 14].”
BackgroundSemi-supervised Sequence Learning
Dai, A. M., & Le, Q. V. (2015) · arXiv · doi:10.48550/arxiv.1511.01432 · Cited by 994
StereoSet: Measuring stereotypical bias in pretrained language models
Nadeem, M., Bethke, A., & Reddy, S. (2020) · arXiv · doi:10.48550/arxiv.2004.09456 · Cited by 643 · cited 2×
“As a result, even though the various language models have different architectures, since they are trained on similar corpora of text, they often learn similar biases [9].”
BackgroundUniversal Adversarial Triggers for Attacking and Analyzing NLP
Wallace, E., Feng, S., Kandpal, N., Gardner, M., & Singh, S. (2019) · Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP) · doi:10.18653/v1/d19-1221 · Cited by 642
Identifying and Reducing Gender Bias in Word-Level Language Models
Bordia, S., & Bowman, S. R. (2019) · Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Student Research Workshop · doi:10.18653/v1/n19-3002 · Cited by 219 · cited 1×
“Previous work has explored the biases that are ultimately learned by language models, with the primary focus being racial and gender stereotypes [2, 3, 7, 12].”
BackgroundThe Woman Worked as a Babysitter: On Biases in Language GenerationPreprint
Sheng, E., Chang, K.-W., Natarajan, P., & Peng, N. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1909.01326 · Cited by 205 · cited 2×
“In [12], the authors made similar observations about language generation models.”
Background
Research works printed but not linked: 6
- 7. T. Bolukbasi, K.-W. Chang, J. Y. Zou, V. Saligrama, A. T. Kalai, Advances in neural information processing systems 29 , 4349 (2016).
- 9. S. Bordia, S. R. Bowman, arXiv preprint arXiv:1904.03035 (2019). · arXiv
- 2. A. M. Dai, Q. V. Le, Advances in neural information processing systems 28 , 3079 (2015).
- 10. K. Lu, P. Mardziel, F. Wu, P. Amancharla, A. Datta, Logic, Language, and Security (Springer, 2020), pp. 189-202.
- 3. A. Vaswani, et al. , Advances in neural information processing systems (2017), pp. 59986008.
- 14. E. Wallace, S. Feng, N. Kandpal, M. Gardner, S. Singh, arXiv preprint arXiv:1908.07125 (2019). · arXiv
Cited by
Items 901–986 of 986
Vu, T., Nguyen, L., & Dao, Q. (2026) · IEEE Access · doi:10.1109/access.2026.3698020
Almeida, J. G. R. T. (2026) · AI & Society · doi:10.1007/s00146-026-03130-6
“…how large language models reproduce and amplify social, cultural, and geographic inequalities embedded in their training data (Bender et al. 2021; Abid et al. 2021; Kerche et al. 2026)— but to demonstrate empirically that the visibility, negotia-bility, and correctability of such bias depend on…”
Toward Generative AI Applications in Disaster Management: A Critical Review and Conceptual Framework
Turan, M. (2026) · Sustainable development · doi:10.5772/intechopen.1014952
Ma, X., Zhu, R., Wang, Z., Xiong, J., Chen, Q., Camp, L. J., & Tang, H. (2026) · Journal of Medical Internet Research · doi:10.2196/89003
“Such biases can surface in outputs and may be amplified under certain prompting strategies, including persona-based approaches [14-16].”
Mohammad, N. I. S., & sheikh, T. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2606.16562
“This design operational-izes the media-coupling mechanism identified by Abid et al. (2021), who observed that anti-Muslim association rises in tandem with terrorism-related news coverage and allows us to isolate its magnitude at inference time as the time-coupling coefficient: where V recent and V…”
InfluentialNo-Free-Fairness: Fundamental Limits and Trade-offs in Learning SystemsPreprint
Than, K. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2606.17810
“Recent empirical studies have documented various forms of bias and inequity in LLMs, including demographic bias [6, 37, 24, 36], representational harms [1, 7], and disparities in toxicity or refusal behavior across social groups [20] or social biases in health care [30].”
DeJesus, A. C. S. (2026) · Journal of International Crisis and Risk Communication Research
A Gender-Diverse Perspective of Bias in Large Language Models
Gaba, A., Wall, E., Keyes, O., Brun, Y., Hall, K. W., & Xiong, C. (2026) · ACM Conference on Fairness, Accountability, and Transparency (FAccT) · doi:10.1145/3805689.3812257
“Recent studies demonstrate that LLMs exhibit biases against religious groups [2], produce gender stereotypes [58], and generate stigmatizing language about non-binary and transgender individuals [46, 48, 53].”
Unearthing implicit biases: a scenario-based evaluation of open-source LLMs
Yao, Y., Hao, L., Rong, H., Dan, W., & Zhao, X. (2026) · Aslib Journal of Information Management · doi:10.1108/ajim-06-2025-0389
Eiger, Y., Hansen, T., Barrett, T., Hutson, J. A., Newell, B. C., & Roesner, F. (2026) · ACM Conference on Fairness, Accountability, and Transparency (FAccT) · doi:10.1145/3805689.3806473
Interkulturális kompetencia - nyelv és mesterséges intelligencia
Gácsi, M. J., & Kiss, A. T. (2026) · Mesterséges intelligencia · doi:10.35406/mi.2026.1.43
Legal Domain Adaptation of Modern BERT ModelsPreprint
Stammbach, D., & Henderson, P. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2606.28538
“LLMs are known to be biased [1, 17], however we did not investigated such biases in our models.”
Liu, D., Sun, H., Wang, Z. G., Li, Q., & Zhang, M. (2026) · Proceedings of the conference - Association for Computational Linguistics. Meeting · doi:10.18653/v1/2026.acl-long.1317
“As LLMs are widely adopted, studies show they often encode societal stereotypes during pretraining (Gallegos et al., 2024a; Ganguli et al., 2023), causing systemic bias even in neutral tasks (Abid et al., 2021b; Wang et al., 2023; Huang et al., 2021).”
Caged Birds and Cute Bookworms: Feminine Tropes and Implicit Gender Bias in Large Language Models
Nishal, S., & Bandy, J. (2026) · Workshop on Evaluating Evaluations (EvalEval) · doi:10.18653/v1/2026.evaleval-1.22
“Studies have shown bi-026 ased representation in terms of gender and occupa-027 tion (Kotek et al., 2023), attitudes towards different religious groups (Abid et al., 2021), heteronorma-029 tive relationships (Gillespie, 2024), descriptions of financial markets (Chuang and Yang, 2022), and more.”
Can We Entrust Justice to AI?: How Persona Traps Contaminate Reasoning in Criminal Investigation
Lee, J., Kang, M.-C., & Shin, J.-H. (2026) · Findings of the Association for Computational Linguistics: ACL (Findings of ACL) · doi:10.18653/v1/2026.findings-acl.843
“These studies have repeatedly confirmed that LLMs absorb and reproduce social biases inherent in training data (Abid et al., 2021; Lee et al., 2024; Hu et al., 2025).”
Sakunkoo, A., & Sakunkoo, J. (2026) · Proceedings of the conference - Association for Computational Linguistics. Meeting · doi:10.18653/v1/2026.acl-srw.96
Detecting AI-Generated Arabic Text Using a Stacking Ensemble of Arabic-Specific Transformers
Daoud, H., Khedr, A. Y., Madkour, M. A., & Elnabwy, M. A. (2026) · ICEENG International Conference for Innovations in Intelligent Computing and Cybersecurity (IICC) · doi:10.1109/iicc69623.2026.11582586
“Synthetic content on digital platforms threatens the accuracy of academic, journalistic, and official communication, and AI-generated Arabic often carries recurring cultural and linguistic biases that further undermine information reliability [5].”
Zhao, X., Zhao, C., Mengesteab, M., Wong, C.-W., Zhan, M. M., Tan, Z., & Chen, T. (2026) · Social Science Computer Review · doi:10.1177/08944393261465001
Mohamed, E. (2026) · Cogent Arts and Humanities · doi:10.1080/23311983.2026.2690890
Li, X., & Yang, Y. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-29903-1_17
MoghadasNian, S., AlizadehMousavi, S., HosseinPoor, S., Farmanian, M., & Bagheri, A. (2026) · doi:10.24108/preprints-3115798
Artificial Intelligence for Bias Detection in Higher Education Online Content
Shiha, R. M. A. B. (2026) · White Rose eTheses Online (University of Leeds, The University of Sheffield, University of York)
Yang, Y., Huang, H., Li, J. F., Sun, H., & Gao, Y. (2026) · ACM Transactions on Information Systems · doi:10.1145/3831687
“Bias in AI systems poses a persistent and multifaceted concern that spans domains from natural language processing and computer vision to recommendation and search systems [1, 6, 12, 20, 21, 27, 33, 34, 38, 49].”
Xia, Y., Batorski, D., Wertz, B., Mamakos, M., Li, L., & Wojcieszak, M. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2607.18931
Chatbots in the role of text generation on social media and their impact on political developments*
Lukinec, M. P., Oreški, D., & Vlahek, D. (2026) · Politička misao : časopis za politologiju · doi:10.20901/pm.63.2.08
Navigating Boundaries of AI Use in Physical Education Journal Research Publications
Li, Y., Wu, C., Su, Y., Lu, B., Hamilton, X. D., Lynch, T., & Hamilton, M. F. (2026) · Quest · doi:10.1080/00336297.2026.2707164
Attia, B. M. (2026) · Journal of Ethnic and Cultural Studies · doi:10.66815/ejecs/2640
“For example, Abid et al. (2021) observed the pervasiveness of anti-Muslim stereotypes in popularly used LLMs (Sultan et al., 2025).”
Red Teaming with Faith Leaders: Expanding Digital Safety and Accountability to Frontiers of CarePreprint
Lutz, N., & Weyl, E. G. (2026) · SSRN Electronic Journal · doi:10.2139/ssrn.6663398
Seeing against the machine: reframing generative AI error as pedagogical insight in art education
Abdulmajid, M. (2026) · Journal of Visual Literacy · doi:10.1080/1051144x.2026.2710031
Asfari, A., Ouassini, N., & Thomsen, E. (2026) · Oxford University Press eBooks · doi:10.1093/9780198945246.003.0137
Coscia, A. J., Duwal, S., Holmes, L., Crossley, S. A., & Endert, A. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2608.04006
Doğan, F. I., Lou, Y., Markelius, A., Geijer‐Simpson, E., Gredebäck, G., Ford, T. J., Castellano, G., Güneş, H., Warner, G., & Gibson, J. L. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2608.03820
“…association between Muslims and violence that surfaces even in story generation from neutral prompts and resists straightforward correction [7]; given that the storytelling and picture-description activities rely on generated or interpreted narrative content, such associations could enter…”
Vance, T. R., & Vance, A. (2026) · ITU Kaleidoscope Academic Conference (ITU K) · doi:10.23919/ituk70038.2026.11632989
“A growing body of empirical research demonstrates that frontier and general purpose models display systematic cultural skew, with disproportionate error and bias rates in non-English languages and underrepresented regional linguistic domains [14][15].”
Hybrid Intelligent Systems: Revisiting Ethical AI Architecture in the Modern Era
Medsker, L., & Virmani, S. (2026) · Springer eBooks · doi:10.1007/978-3-032-12201-8_14
Natural Language Inference of Moral Change
Ramezani, A. (2026) · TSpace (University of Toronto)
Beyond Surface Cues: Disentangling Sociocultural Signals in Multilingual LLMsPreprint
Feng, Y., Liu, T., Feuerriegel, S., & Shrestha, Y. R. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2608.23026
“The influence of language structure on the reproduction of gender stereotypes is deeply connected to cultural context and the composition of training data (Abid et al., 2021; Zhong et al., 2024).”
BEYOND FAIRNESS METRICS: HUMAN EXPERIENCES AND DIFFERENTIAL OUTCOMES IN AI-MEDIATED WORK SETTINGS
Kadoma, K. (2026) · eCommons (Cornell University) · doi:10.7298/4h41-w092
Visuri, I., & Svensson, J. (2026) · Religion & Education · doi:10.1080/15507394.2026.2716456
“The problem of stereotypes, particularly in the context of representing religious positions, is real and has been the object of empirical research (see e.g. Abid et al., 2021; Plaza-del-Arco et al., 2024).”
Exploring Fairness and Bias Mitigation in Large Language Models
Jaidi, C. R. (2025) · NORMA
Is Your Neighborhood Safe? Place-based Stigma in Large Language Models' Urban Safety JudgmentsPreprint
Nguyen, H., & Lin, Y. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2608.26188
“More recent work extends this to bias and stereotype in modern language generation and large models [1, 20, 28], with broad analyses of scale warning that fluent models can encode and amplify social stereotypes [2, 29].”
Hung, H. T., Midha, N., Wu, V. Y., & Zhang, Y. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2609.04243
Mapping the Emerging Social Science of Large Language ModelsPreprint
Zhao, Z., Yang, Y., Jia, X., Dong, Z., & Wang, C. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2609.07598
Exploring Backdoor Vulnerabilities and Fairness Challenges in Generative Models: A Survey
Holland, R., Pal, S., & Pan, L. (2026) · ACM Computing Surveys · doi:10.1145/3847107
When algorithms discriminate: AI bias and its implications for tourism and hospitality
Solarević, M., Kalinić, Č., Vujičić, M. D., & Stankov, U. (2026) · Zbornik radova Departmana za geografiju turizam i hotelijerstvo · doi:10.5937/zbdght2601096s
“Since tourism is global, biases may have different impacts depending on the cultural context and the digital liter - acy of users (Abid et al., 2021).”
From policy to practice: clause‑grounded answers with retrieval‑augmented generation
Tsai, W.-L., & Huang, R.-Q. (2026) · Journal of Experimental & Theoretical Artificial Intelligence · doi:10.1080/0952813x.2026.2729304
When Who You Are Can Change the Code You Get: A Study of Persona-Induced Bias in LLM Code GenerationPreprint
Gupta, A., Figueiredo, M. C., Machado, L. S., Wright, T., Beschastnikh, I., de Souza, C. R. B., & Rodríguez-Pérez, G. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2609.22102
“Similar forms of bias have been identified across other dimensions, including religion [1], disability [54], and sexuality [17].”
LangBiTe: model-driven bias testing of text-to-text large language models
Morales, S., Clarisó, R., & Cabot, J. (2026) · Software & Systems Modeling · doi:10.1007/s10270-026-01425-2
Access denied : paywalls, information inequality, and the economics of quality journalism
CARLINI, R., Borges, D. D. C. L., González, N. M., & Mazzoni, L. (2026) · Lirias · doi:10.2870/5239572
Alperen, A. V. C. I., & Demir, E. (2026) · KADEM Kadın Araştırmaları Dergisi · doi:10.21798/kadem.2026.220
Poppi, T., Cappelletti, S., Poppi, S., Cornia, M., Baraldi, L., García-Olano, D., & Cucchiara, R. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2609.39688
“These models derive their generalization abilities from massive web-scale datasets, but uncontrolled data collection can introduce harmful biases (Birhane, Prabhu, & Kahembwe, 2021), unsafe associations (Abid, Farooqi, & Zou, 2021), and inappropriate content (Schuhmann et al., 2022).”
Corciulo, S. (n.d.) · Institutional Research Information System University of Turin (University of Turin)
Measuring and mitigating stereotypes in large language models
Siddique, Z. (2025) · ORCA Online Research @Cardiff (Cardiff University)
Basyar, S. (2026) · Advances In Education Journal
THE IMPACT OF CHATGPT ON MODERN LINGUISTICS AND THE CHALLENGES IT POSES
Dilshodovna, N. Z. (2025) · Modern American Journal of Linguistics, Education, and Pedagogy
Torres-Toukoumidis, Á., León-Alberca, T., & Lituma, J. P. C. (2026) · Current Psychology · doi:10.1007/s12144-026-10121-9
“Moderation sys - tems have been reported to conflate the display of higher-weight bodies with sexualised content, reducing the reach of fat activist accounts (Abid et al., 2021 ; Álvarez-Castillo, 2015; Del Mar et al., 2025).”
Molecules of a story:Community detection in narrative networks to unearth micro-narratives
Fyhn, K., & Baglini, R. (2026)
Yang, L., & Lyu, Y.-N. (2026) · International Conference on Information Technology for Social Good (GoodIT) · doi:10.1145/3794786.3830740
Group theory and information theory algorithms in deep learning
Basu, S. (2024) · Illinois Digital Environment for Access to Learning and Scholarship (University of Illinois at Urbana-Champaign)
Bias in Large Language Models Through the Lens of Data Feminism
Wang, L. (2026) · Applied and Computational Engineering · doi:10.54254/2755-2721/2026.gf37475
Understanding and Addressing AI Vulnerability and Digital Gaps in IsraelPreprint
Sierra, M. (2026) · SSRN Electronic Journal · doi:10.2139/ssrn.7569398
Intergroup Conflict and Human-Algorithm Interactions
Kyrychenko, Y. (2026) · Apollo (University of Cambridge) · doi:10.17863/cam.135589
ClearDose: When Personalized Human-AI Interactions Are Never Just Personal
Han, Y., cai, S., Li, X., & Leigh, S.-W. (2026) · ACM Conference on Computer Supported Cooperative Work (CSCW) · doi:10.1145/3785651.3831417
Localized, unstable, and entangled: Exploring `Us vs. Them' bias in Large Language ModelsPreprint
Prama, T. T., Zimmerman, J. W., Danforth, C. M., & Dodds, P. S. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2610.10864
“Bias in LLMs. Prior work shows that LLMs reproduce social biases related to gender, race, religion, nationality, age, social identity, political, and other demographic attributes [9, 36–40].”
Rechtliche Aspekte des qualitativen Forschens zusammen mit KI
Schäffer, B., Lieder, F. R., & Krämer, F. (2026) · Springer eBooks · doi:10.1007/978-3-658-53148-5_7
The Implications of Large Language Model Integration for Bias in Lending Practices
Flores-Torres, D., Lee, R. L., & Blum, J. J. (2026) · Communications in computer and information science · doi:10.1007/978-3-032-28469-3_18
Responsible AI and the Middle Ages: Detecting Historical Toxicity in Medieval Datasets
Demelas, D. (2026) · Digital humanities quarterly · doi:10.63744/ed7kpzmuudb2
Who Do Perspectivist Models Predict Well? Annotator Agreement, Sparsity, and PerformancePreprint
Lowmanstone, L., Sivakumar, V., Samaddar, A., Mundada, A., Kaur, H., & Gini, M. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2610.22567
“Other work attempts to measure bias in existing NLP systems (Nangia et al., 2020; Parrish et al., 2022; Nadeem et al., 2021; Abid et al., 2021; Bolukbasi et al., 2016).”
FAIRITHM: Unlearning Bias and Rewriting for Fairness
Nam, Y., Lim, D., Im, S., Xiong, Z., & Park, J. (2026) · International Journal of Human-Computer Interaction · doi:10.1080/10447318.2026.2744493
What is "Cultural Representation” in AI Benchmarks? A Multidimensional Evaluation Framework
Holtermann, C., Birkmann, J., & Lauscher, A. (2026) · doi:10.18653/2026.pandora-1.3
Algoritmik İslamofobi: Yapay Zekâ Sistemlerinde Müslüman Kimliğinin Üç Düzeyli Yeniden Üretimi
Kaya, E. E. (2026) · İçtimaiyat · doi:10.33709/ictimaiyat.1984784
Artificial Intelligence and Religious Policy: Ethical Governance in the Digital Public Sphere
Hakim, W., Huda, N., & Thang, T. T. (2026) · AMUYA INDONESIAN JOURNAL OF MANAGEMENT REVIEWS · doi:10.61860/amuya.v2i2.122
Nushra, S., & Morol, M. K. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2610.40996
“Most factuality benchmarks are English-first [1, 2], and anti-Muslim bias has been documented in language models [3].”
Qistas: A Pre-Registered, Portable Protocol for Fairness Evaluation of Large Language ModelsPreprint
Alseelawi, N. S., & Aljumaily, M. S. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2610.40893
“…many axes and Multilingual HolisticBias extends them to fifty languages [6]; DiscrimEval [26] probes decision scenarios with demographic substitutions; and Abid et al. [1] showed, early and with a single religious group, that a completion model’s associations can be measured with matched prompts.”
Pride in Precision, Prejudice in Prediction: Bias and Stereotyping in Recommendation Systems
Kheya, T. A. (2026) · ACM International Conference on Information and Knowledge Management (CIKM) · doi:10.1145/3799682.3840226
Souza, D., Noce, H. L., Farias, V. A. E., Machado, J. C., & Caminha, C. (2026) · International Conference on Machine Learning and Applications · doi:10.1109/icmla69890.2026.00214
A era da decisão assistida: o papel da inteligência artificial generativa
Rolemberg, T. M. (2026) · TECCOGS Revista Digital de Tecnologias Cognitivas · doi:10.23925/1984-3585.2026i33p53-72
FAIR: Balancing Identity-Associated Training Signals to Mitigate Personalization Bias
Vijjini, A. R., & Chaturvedi, S. (2026) · doi:10.18653/2026.findings-aacl.160
CompAIssionate Other: Identity Assumptions in LLM Responses to DistressPreprint
Goel, D., Nguyen, E. M., Rodriguez, V. J., & Saha, K. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2611.10808
Bojić, L., Olenik‐Shemesh, D., Ergüleç, F., Rey, R. D., Norman, J. O., Heiman, T., Mora-Merchán, J. A., Meehan, A., Uçar, S., & Kunić, T. (2026) · IEEE International Symposium on Systems Engineering · doi:10.1109/isse70179.2026.11762949
Governing religious outputs of generative AI: a maqāṣid -based framework in light of fiqh al-maʾālāt
Samoh, A., & Samoh, H. (2026) · Cogent Arts and Humanities · doi:10.1080/23311983.2026.2759439
Al-Azri, S. (2026) · IGI Global eBooks · doi:10.4018/979-8-3373-7421-5.ch015
Large Language Models are Geographically Biased
Manvi, R., Khanna, S., Burke, M., Lobell, D., & Ermon, S. (2024) · arXiv · doi:10.48550/arxiv.2402.02680
“…et al., 2020; Garimella et al., 2021; Nangia et al., 2020; Tan & Celis, 2019), ethnic bias (Ahn & Oh, 2021; Garg et al., 2017; Li et al., 2020; Abid et al., 2021; Manzini et al., 2019; Venkit et al., 2023), age bias (Nangia et al., 2020; Diaz et al., 2018), and sexual-orientation bias (Nangia…”
BackgroundUnderstanding Generative AI Risks for Youth: A Taxonomy Based on Empirical Data
Yu, Y., Liu, Y., Zhang, J., Huang, Y., & Wang, Y. (2025) · arXiv · doi:10.48550/arxiv.2502.16383
“Existing research has highlighted that AI systems, particularly LLMs, can perpetuate bias and discrimination through forms including misrepresentation, stereotyping, disparate performance, derogatory language, and exclusionary norms [3, 16, 18, 40].”
Metadata
- Authors
- Abid, A., Farooqi, M., & Zou, J.
- Affiliations
- Stanford University (US)
- McMaster University (CA)
- Venue
- Proceedings of the 2021 AAAI/ACM Conference on AI, Ethics, and Society
- Publisher
- ACM
- Published
- 2021-07-21
- Pages
- 298-306
- ISSN
- 3065-8365 (ISSN-L)
- OpenAlex
- W3184144760
- Cited by
- 476
- References
- 3
- Citations by year
- 2026: 81 · 2025: 124 · 2024: 105 · 2023: 97 · 2022: 36 · 2021: 33
- FWCI
- 61.36
- Type
- conference paper
- Language
- en
- Open access
- green
- arxiv.org
- Locations
- Proceedings of the AAAI/ACM Conference on AI Ethics and Society (publishedVersion)
- arXiv (Cornell University) (submittedVersion, cc-by)
- arXiv (Cornell University) (cc-by)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Ethics and Social Impacts of AI
- Physical Sciences › Computer Science › Artificial Intelligence › Hate Speech and Cyberbullying Detection
- Physical Sciences › Computer Science › Artificial Intelligence › Topic Modeling
- Keywords
- Islamophobia, religious bias, large language models, GPT-3, analogical reasoning, adversarial prompts, story generation, bias mitigation
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.