Revisão Rápida sobre Vieses em Chatbots - Uma análise sobre tipos de vieses, impactos e formas de lidar

  • Thiago M. R. Ribeiro Universidade Federal do Estado do Rio de Janeiro (UNIRIO)
  • Sean W. M. Siqueira Universidade Federal do Estado do Rio de Janeiro (UNIRIO)
  • Maira G. de Bayser Universidade Federal do Estado do Rio de Janeiro (UNIRIO)


Devido ao seu funcionamento, chatbots podem perpetuar vieses cognitivos e sociais, cujos impactos precisam ser avaliados. Foi realizada uma revisão rápida, contemplando entrevista e grupo focal de especialistas em Tecnologia da Informação e Comunicação, além de uma busca na base SCOPUS, para identificar na literatura os impactos dos vieses em chatbots. De 488 estudos encontrados, foram selecionados 18 para a análise final. Ao todo, sete tipos de vieses diferentes emergiram dos estudos, assim como os seus impactos positivos e negativos, seus domínios e formas de mitigação. A contribuição esperada com este estudo consiste no aprimoramento de ferramentas conversacionais, bem como apoiar os usuários na identificação e mitigação de vieses.

Palavras-chave: revisão rápida, vieses, llm, chatbots, sistemas conversacionais, ChatGPT


Abid, A., Farooqi, M., and Zou, J. (2021). Persistent anti-muslim bias in large language models. In Proceedings of the 2021 AAAI/ACM Conference on AI, Ethics, and Society, pages 298–306.

Adamopoulou, E. and Moussiades, L. (2020). An overview of chatbot technology. In IFIP international conference on artificial intelligence applications and innovations, pages 373–383.

Bakshy, E., Messing, S., and Adamic, L. A. (2015). Exposure to ideologically diverse news and opinion on facebook. Science, 348(6239):1130–1132.

Bansal, H. and Khan, R. (2018). A review paper on human computer interaction. Int. J. Adv. Res. Comput. Sci. Softw. Eng, 8(4):53.

Basili, V. R. and Rombach, H. D. (1988). The tame project: Towards improvement-oriented software environments. IEEE Transactions on software engineering, 14(6):758–773.

Beattie, H., Watkins, L., Robinson, W. H., Rubin, A., and Watkins, S. (2022). Measuring and mitigating bias in ai-chatbots. In 2022 IEEE International Conference on Assured Autonomy (ICAA), pages 117–123. IEEE.

Bessi, A., Zollo, F., Del Vicario, M., Puliga, M., Scala, A., Caldarelli, G., Uzzi, B., and Quattrociocchi, W. (2016). Users polarization on facebook and youtube. PloS one, 11(8):e0159641.

Bonaccio, S. and Dalal, R. S. (2006). Advice taking and decision-making: An integrative literature review, and implications for the organizational sciences. Organizational behavior and human decision processes, 101(2):127–151.

Bouchet, F. and Sansonnet, J.-P. (2009). Subjectivity and cognitive biases modeling for a realistic and efficient assisting conversational agent. In 2009 IEEE/WIC/ACM International Joint Conference on Web Intelligence and Intelligent Agent Technology, volume 2, pages 209–216. IEEE.

Cartaxo, B., Pinto, G., and Soares, S. (2018). The role of rapid reviews in supporting decision-making in software engineering practice. In Proceedings of the 22nd International Conference on Evaluation and Assessment in Software Engineering 2018, pages 24–34.

Felkner, V. K., Chang, H.-C. H., Jang, E., and May, J. (2023). Winoqueer: A community in-the-loop benchmark for anti-lgbtq+ bias in large language models. arXiv preprint arXiv:2306.15087.

Ferrara, E. (2023). Should chatgpt be biased? challenges and risks of bias in large language models. arXiv preprint arXiv:2304.03738.

Fossa, F. and Sucameli, I. (2022). Gender bias and conversational agents: an ethical perspective on social robotics. Science and Engineering Ethics, 28(3):23.

Garrido-Munoz, I., Martínez-Santiago, F., and Montejo-Raez, A. (2023). Maria and beto are sexist: evaluating gender bias in large language models for spanish. Language Resources and Evaluation, pages 1–31.

Ghosh, S. and Caliskan, A. (2023). Chatgpt perpetuates gender bias in machine translation and ignores non-gendered pronouns: Findings across bengali and five other low-resource languages. arXiv preprint arXiv:2305.10510.

Gross, N. (2023). What chatgpt tells us about gender: A cautionary tale about performativity and gender biases in ai. Social Sciences, 12(8):435.

Hayashi, Y., Takii, S., Nakae, R., and Ogawa, H. (2012). Exploring egocentric biases in human cognition: An analysis using multiple conversational agents. In 2012 IEEE 11th International Conference on Cognitive Informatics and Cognitive Computing, pages 289–294. IEEE.

Kahneman, D. and Tversky, A. (1972). Subjective probability: A judgment of representativeness. Cognitive psychology, 3(3):430–454.

Kempf, A. (2020). If we are going to talk about implicit race bias, we need to talk about structural racism: Moving beyond ubiquity and inevitability in teaching and learning about race. Taboo: The Journal of Culture and Education, 19(2):10.

Khanna, A., Pandey, B., Vashishta, K., Kalia, K., Pradeepkumar, B., and Das, T. (2015). A study of today’s ai through chatbots and rediscovery of machine intelligence. International Journal of u-and e-Service, Science and Technology, 8(7):277–284.

Kolisko, S. and Anderson, C. J. (2023). Exploring social biases of large language models in a college artificial intelligence course.

Lee, H., Hong, S., Park, J., Kim, T., Kim, G., and Ha, J.-W. (2023). Kosbi: A dataset for mitigating social bias risks towards safer large language model application. arXiv preprint arXiv:2305.17701.

Motoki, F., Neto, V. P., and Rodrigues, V. (2023). More human than human: Measuring chatgpt political bias. Public Choice, pages 1–21.

Navigli, R., Conia, S., and Ross, B. (2023). Biases in large language models: Origins, inventory and discussion. ACM Journal of Data and Information Quality, 15(2):1–21.

Neff, G. (2016). Talking to bots: Symbiotic agency and the case of tay. International Journal of Communication.

Nozza, D., Bianchi, F., Hovy, D., et al. (2022). Pipelines for social bias testing of large language models. In Proceedings of BigScience Episode# 5–Workshop on Challenges & Perspectives in Creating Large Language Models. Association for Computational Linguistics.

Pfeuffer, N., Adam, M., Toutaoui, J., Hinz, O., and Benlian, A. (2019). Mr. and mrs. conversational agent-gender stereotyping in judge-advisor systems and the role of egocentric bias.

Pizzol, S. J. S. (2004). Combinação de grupos focais e análise discriminante: um método para tipificação de sistemas de produção agropecuária. Revista de Economia e Sociologia Rura, 42:451–468.

Pohl, R. and Pohl, R. F. (2004). Cognitive illusions: A handbook on fallacies and biases in thinking, judgement and memory. Psychology Press.

Ray, P. P. (2023). Chatgpt: A comprehensive review on background, applications, key challenges, bias, ethics, limitations and future scope. Internet of Things and Cyber-Physical Systems, 3:121–154.

Ribeiro, B. B., Resende, J. A., Ribeiro, T. M. R., Santos, R. P., and Siqueira, S. W. M. (2023). Mapeamento sistemático sobre vieses cognitivos no desenvolvimento de software. In Anais do VIII Workshop sobre Aspectos Sociais, Humanos e Economicos de Software, pages 21–30. SBC.

Rozado, D. (2023). The political biases of chatgpt. Social Sciences, 12(3):148. Rufino Junior, R., Classe, T. M., and Santos, R. P. (2022). Jogos digitais para treinamento de situações de risco na indústria - rapid review. In Anais Estendidos do XXI Simposio Brasileiro de Jogos e Entretenimento Digital, pages 1157–1166.

Santhanam, S., Karduni, A., and Shaikh, S. (2020). Studying the effects of cognitive biases in evaluation of conversational agents. In Proceedings of the 2020 CHI Conference on Human Factors in Computing Systems, pages 1–13.

Shawar, B. A. and Atwell, E. (2007). Chatbots: are they really useful? Journal for Language Technology and Computational Linguistics, 22(1):29–49.

Smith, G. and Rustagi, I. (2021). When good algorithms go sexist: Why and how to advance ai gender equity. Standford Social Innovation Review.

Treude, C. and Hata, H. (2023). She elicits requirements and he tests: Software engineering gender bias in large language models. arXiv preprint arXiv:2303.10131.

Wallace, E., Feng, S., Kandpal, N., Gardner, M., and Singh, S. (2019). Universal adversarial triggers for attacking and analyzing nlp. arXiv preprint arXiv:1908.07125.

Yeon, J., Park, Y., and Kim, D. (2023). Is gender-neutral ai the correct solution to gender bias. Using Speech.

Zemčík, T. (2021). Failure of chatbot tay was evil, ugliness and uselessness in its nature or do we judge it through cognitive shortcuts and biases? AI & SOCIETY, 36:361–367.
RIBEIRO, Thiago M. R.; SIQUEIRA, Sean W. M.; DE BAYSER, Maira G.. Revisão Rápida sobre Vieses em Chatbots - Uma análise sobre tipos de vieses, impactos e formas de lidar. In: SIMPÓSIO BRASILEIRO DE SISTEMAS COLABORATIVOS (SBSC), 19. , 2024, Salvador/BA. Anais [...]. Porto Alegre: Sociedade Brasileira de Computação, 2024 . p. 56-70. ISSN 2326-2842. DOI: