Who gets represented by Indonesian AI? measuring regional, gender, and sociolinguistic bias in large language models

Mohammad Anis Sumadi

doi:10.67490/ijcl.v1i2.917

Authors

Mohammad Anis Sumadi Institut Agama Islam Al-Urwatul Wutsqo Author

DOI:

https://doi.org/10.67490/ijcl.v1i2.917

Keywords:

algorithmic bias, gender representation, Indonesian AI, large language models, sociolinguistic bias

Abstract

Background: Indonesia’s regional, gendered, and sociolinguistic diversity raises a critical question about whether large language models represent Indonesian identities with equal specificity, agency, and legitimacy in computational discourse. Objective: This study aims to examine how Indonesian-facing large language models generate representations of regions, gender markers, occupations, and language varieties under controlled prompt conditions. Method: Using a prompt-based audit design, this study analyses 42 prompt units divided into regional, gender-counterfactual, and sociolinguistic conditions, with coding focused on visibility, specificity, agency, competence, register alignment, semantic stability, and language shifting. Results: The findings indicate that regional representation is uneven: some regions are profiled through professional competence, while others are rendered through generic neutrality, cultural tokenisation, peripheral framing, or national homogenisation. Gendered outputs show partial professional parity, but male-coded subjects receive stronger leadership and technical authority, whereas female-coded subjects are more often associated with care, affect, and relational labour. Implication: Sociolinguistic robustness is strongest in formal Indonesian, more adaptive in colloquial Indonesian, and less stable in local-language conditions, where semantic drift, code-mixing, and defaulting to Indonesian appear. Novelty: This study contributes an intersectional audit framework that reframes Indonesian AI bias as a problem of regional visibility, gendered agency, and sociolinguistic legitimacy across culturally stratified AI systems in multilingual Indonesia today.

References

[1] X. Bai, A. Wang, I. Sucholutsky, and T. L. Griffiths, “Explicitly unbiased large language models still form biased associations,” Proceedings of the National Academy of Sciences of the United States of America, vol. 122, 2025, doi: 10.1073/pnas.2416228122.

[2] H. Kotek, R. Dockum, and D. Q. Sun, “Gender bias and stereotypes in large language models,” Proceedings of the ACM Collective Intelligence Conference, 2023, doi: 10.1145/3582269.3615599.

[3] M. Kaneko, A. Imankulova, D. Bollegala, and N. Okazaki, “Gender bias in masked language models for multiple languages,” arXiv, 2022, doi: 10.48550/arxiv.2205.00551.

[4] J. Zhao, Y. Ding, C. Jia, Y. Wang, and Z. Qian, “Gender bias in large language models across multiple languages,” arXiv, 2024, doi: 10.48550/arxiv.2403.00277.

[5] R. Manvi, S. Khanna, M. Burke, D. B. Lobell, and S. Ermon, “Large language models are geographically biased,” arXiv, 34654–34669, 2024, doi: 10.48550/arxiv.2402.02680.

[6] T. Naous, M. J. Ryan, and W. Xu, “Having beer after prayer? Measuring cultural bias in large language models,” arXiv, 16366–16393, 2023, doi: 10.48550/arxiv.2305.14456.

[7] Y. Tao, O. Viberg, R. S. Baker, and R. F. Kizilcec, “Cultural bias and cultural alignment of large language models,” PNAS Nexus, vol. 3, 2023, doi: 10.1093/pnasnexus/pgae346.

[8] L. C. L. Gamboa and M. Lee, “Filipino benchmarks for measuring sexist and homophobic bias in multilingual language models from Southeast Asia,” arXiv, 2024, doi: 10.48550/arxiv.2412.07303.

[9] A. Abid, M. Farooqi, and J. Y. Zou, “Persistent anti-Muslim bias in large language models,” Proceedings of the 2021 AAAI/ACM Conference on AI, Ethics, and Society, 2021, doi: 10.1145/3461702.3462624.

[10] I. O. Gallegos, R. A. Rossi, J. Barrow, M. M. Tanjim, S. Kim, F. Dernoncourt, et al., “Bias and fairness in large language models: A survey,” Computational Linguistics, vol. 50, pp. 1097–1179, 2023, doi: 10.1162/coli_a_00524.

[11] R. Navigli, S. Conia, and B. Ross, “Biases in large language models: Origins, inventory, and discussion,” ACM Journal of Data and Information Quality, vol. 15, pp. 1–21, 2023, doi: 10.1145/3597307.

[12] E. Ferrara, “Should ChatGPT be biased? Challenges and risks of bias in large language models,” First Monday, vol. 28, 2023, doi: 10.5210/fm.v28i11.13346.

[13] I. A. Hanif, M. F. Azmi, F. A. Tjiaranata, E. P. Yulianrifat, and F. Koto, “IndoBias: A dual track culturally grounded benchmark for LLMs bias evaluation in Indonesian languages,” , 2026.

[14] E. Durmus, K. Nyugen, T. Liao, N. Schiefer, A. Askell, A. Bakhtin, et al., “Towards measuring the representation of subjective global opinions in language models,” arXiv, 2023.

[15] Y. Xu, L. Hu, J. Zhao, Z. Qiu, Y. Ye, and H. Gu, “A survey on multilingual large language models: Corpora, alignment, and bias,” Frontiers of Computer Science, vol. 19, 2024, doi: 10.1007/s11704-024-40579-4.

[16] Y. Guo, M. Guo, J. Su, Z. Yang, M. Zhu, H. Li, M. Qiu, and S. Liu, “Bias in large language models: Origin, evaluation, and mitigation,” arXiv, 2024, doi: 10.3390/electronics15091824.

[17] A. Caliskan, “Artificial intelligence, bias, and ethics,” Proceedings of the Thirty-Second International Joint Conference on Artificial Intelligence, 7007–7013, 2023, doi: 10.24963/ijcai.2023/799.

[18] P. Liang, C. Wu, L.-P. Morency, and R. Salakhutdinov, “Towards understanding and mitigating social biases in language models,” Proceedings of the 38th International Conference on Machine Learning, 6565–6576, 2021.

[19] F. Kazi, A. Young, Y. Inani, and S. Rafatirad, “A comprehensive study of implicit and explicit biases in large language models,” arXiv, 2025, doi: 10.48550/arxiv.2511.14153.

[20] J. Mickel, M. De-Arteaga, L. Liu, and K. Tian, “More of the same: Persistent representational harms under increased representation,” arXiv, 2025, doi: 10.48550/arxiv.2503.00333.

[21] Y. Li, G. Zhang, B. Yang, C. Lin, S. Wang, A. Ragni, and J. Fu, “HERB: Measuring hierarchical regional bias in pre-trained language models,” arXiv, 2022, doi: 10.48550/arxiv.2211.02882.

[22] J. An, D. Huang, C. Lin, and M. Tai, “Measuring gender and racial biases in large language models: Intersectional evidence from automated resume evaluation,” PNAS Nexus, vol. 4, 2025, doi: 10.1093/pnasnexus/pgaf089.

[23] A. Haim, A. Salinas, and J. Nyarko, “What’s in a name? Auditing large language models for race and gender bias,” arXiv, 2024, doi: 10.48550/arxiv.2402.14875.

[24] A. Mandal, S. Little, and S. Leavy, “Gender bias in multimodal models: A transnational feminist approach considering geographical region and culture,” arXiv, 2023, doi: 10.48550/arxiv.2309.04997.

[25] L. Jiang, G. Zhu, J. Sun, J. Cao, and J. Wu, “Exploring the occupational biases and stereotypes of Chinese large language models,” Scientific Reports, vol. 15, 2025, doi: 10.1038/s41598-025-03893-w.

[26] S. Levy, N. A. John, L. Liu, Y. Vyas, Y. Fujinuma, M. Ballesteros, V. Castelli, and D. Roth, “Comparing biases and the impact of multilingual training across multiple languages,” arXiv, 10260–10280, 2023, doi: 10.48550/arxiv.2305.11242.

[27] N. R. Sahoo, P. Kulkarni, N. Asad, A. Ahmad, T. Goyal, A. Garimella, and P. Bhattacharyya, “IndiBias: A benchmark dataset to measure social biases in language models for Indian context,” arXiv, 8786–8806, 2024, doi: 10.48550/arxiv.2403.20147.

[28] K. Khandelwal, M. Tonneau, A. M. Bean, H. R. Kirk, and S. A. Hale, “Indian-BhED: A dataset for measuring India-centric biases in large language models,” Proceedings of the 2024 International Conference on Information Technology for Social Good, 2023, doi: 10.1145/3677525.3678666.

[29] A. Vashishtha, K. Ahuja, and S. Sitaram, “On evaluating and mitigating gender biases in multilingual settings,” arXiv, 2023, doi: 10.48550/arxiv.2307.01503.