Academia.eduAcademia.edu

Outline

Title
Abstract
Introduction
Xai Approaches as Questions
Questioning Xai Approaches
From Consequential to Constructive Recommendations
Recommendations for Ai Policy and Governance
Conclusion
References
All Topics
Computer Science
Artificial Intelligence

A Genealogical Approach to Algorithmic Bias

Profile image of David WatsonDavid Watson

2024, Minds and Machines

https://doi.org/10.1007/S11023-024-09672-2
visibility

…

description

17 pages

link

1 file

Abstract

The Fairness, Accountability, and Transparency (FAccT) literature tends to focus on bias as a problem that requires ex post solutions (e.g. fairness metrics), rather than addressing the underlying social and technical conditions that (re)produce it. In this article, we propose a complementary strategy that uses genealogy as a constructive, epistemic critique to explain algorithmic bias in terms of the conditions that enable it. We focus on XAI feature attributions (Shapley values) and counterfactual approaches as potential tools to gauge these conditions and offer two main contributions. One is constructive: we develop a theoretical framework to classify these approaches according to their relevance for bias as evidence of social disparities. We draw on Pearl's ladder of causation (Causality: models, reasoning, and inference.

Related papers

Bias – A Lurking Danger that Can Convert Algorithmic Systems into Discriminatory Entities Towards a Framework for Bias Identification and Mitigation
Eduard Klein

2019

Bias in algorithmic systems is a major cause of unfair and discriminatory decisions in the use of such systems. Cognitive bias is very likely to be reflected in algorithmic systems as humankind aims to map Human Intelligence (HI) to Artificial Intelligence (AI). An extensive literature review on the identification and mitigation of bias leads to precise measures for project teams building AI-systems. Aspects like AI-responsibility, AI-fairness and AI-safety are addressed by developing a framework that can be used as a guideline for project teams. It proposes measures in the form of checklists to identify and mitigate bias in algorithmic systems considering all steps during system design, implementation and

View PDFchevron_right
Bias – A Lurking Danger that Can Convert Algorithmic Systems into Discriminatory Entities
Eduard Klein

2020

Bias in algorithmic systems is a major cause of unfair and discriminatory decisions in the use of such systems. Cognitive bias is very likely to be reflected in algorithmic systems as humankind aims to map Human Intelligence (HI) to Artificial Intelligence (AI). An extensive literature review on the identification and mitigation of bias leads to precise measures for project teams building AI-systems. Aspects like AI-responsibility, AI-fairness and AI-safety are addressed by developing a framework that can be used as a guideline for project teams. It proposes measures in the form of checklists to identify and mitigate bias in algorithmic systems considering all steps during system design, implementation and application.

View PDFchevron_right
Algorithmic Fairness and Bias in Machine Learning Systems
Mr. Karun Sanjaya

E3S web of conferences, 2023

In recent years, research into and concern over algorithmic fairness and bias in machine learning systems has grown significantly. It is vital to make sure that these systems are fair, impartial, and do not support discrimination or social injustices since machine learning algorithms are becoming more and more prevalent in decision-making processes across a variety of disciplines. This abstract gives a general explanation of the idea of algorithmic fairness, the difficulties posed by bias in machine learning systems, and different solutions to these problems. Algorithmic bias and fairness in machine learning systems are crucial issues in this regard that demand the attention of academics, practitioners, and policymakers. Building fair and unbiased machine learning systems that uphold equality and prevent discrimination requires addressing biases in training data, creating fairnessaware algorithms, encouraging transparency and interpretability, and encouraging diversity and inclusivity.

View PDFchevron_right
Towards a critical race methodology in algorithmic fairness
Jamila Loud

Proceedings of the 2020 Conference on Fairness, Accountability, and Transparency, 2020

We examine the way race and racial categories are adopted in algorithmic fairness frameworks. Current methodologies fail to adequately account for the socially constructed nature of race, instead adopting a conceptualization of race as a fixed attribute. Treating race as an attribute, rather than a structural, institutional, and relational phenomenon, can serve to minimize the structural aspects of algorithmic unfairness. In this work, we focus on the history of racial categories and turn to critical race theory and sociological work on race and ethnicity to ground conceptualizations of race for fairness research, drawing on lessons from public health, biomedical research, and social survey research. We argue that algorithmic fairness researchers need to take into account the multidimensionality of race, take seriously the processes of conceptualizing and operationalizing race, focus on social processes which produce racial inequality, and consider perspectives of those most affected by sociotechnical systems.

View PDFchevron_right
The Man Behind the Curtain: Appropriating Fairness in AI
Emanuele Martinelli

Minds and machines, 2024

Our goal in this paper is to establish a set of criteria for understanding the meaning and sources of attributing (un)fairness to AI algorithms. To do so, we first establish that (un)fairness, like other normative notions, can be understood in a proper primary sense and in secondary senses derived by analogy. We argue that AI algorithms cannot be said to be (un)fair in the proper sense due to a set of criteria related to normativity and agency. However, we demonstrate how and why AI algorithms can be qualified as (un)fair by analogy and explore the sources of this (un)fairness and the associated problems of responsibility assignment. We conclude that more user-driven AI approaches could alleviate some of these difficulties.

View PDFchevron_right
Framework for developing algorithmic fairness
Harco Leslie Hendric Spits Warnars

Bulletin of Electrical Engineering and Informatics

In a world where the algorithm can control the lives of society, it is not surprising that specific complications in determining the fairness in the algorithmic decision will arise at some point. Machine learning has been the de facto tool to forecast a problem that humans cannot reliably predict without injecting some amount of subjectivity in it (i.e., eliminating the “irrational” nature of humans). In this paper, we proposed a framework for defining a fair algorithm metric by compiling information and propositions from various papers into a single summarized list of fairness requirements (guideline alike). The researcher can then adopt it as a foundation or reference to aid them in developing their interpretation of algorithmic fairness. Therefore, future work for this domain would have a more straightforward development process. We also found while structuring this framework that to develop a concept of fairness that everyone can accept, it would require collaboration with other...

View PDFchevron_right
Bias and Discrimination in AI: A Cross-Disciplinary Perspective
Mark Cote

IEEE Technology and Society Magazine

 Operating at a large scale and impacting large groups of people, automated systems can make consequential and sometimes contestable decisions. Automated decisions can impact a range of phenomena, from credit scores to insurance payouts to health evaluations. These forms of automation can become problematic when they place certain groups or people at a systematic disadvantage. These are cases of discrimination-which is legally defined as the unfair or unequal treatment of an individual (or group) based on certain protected characteristics (also known as protected attributes) such as income, education, gender, or ethnicity. When the unfair treatment is caused by automated decisions, usually taken by intelligent agents or other AI-based systems, the topic of digital discrimination arises. Digital discrimination is prevalent in a diverse range of fields, such as in risk assessment systems for policing and credit scores [1], [2]. Digital discrimination is becoming a serious problem, as more and more decisions are delegated to systems increasingly based on artificial intelligence (AI) techniques such as machine learning. Although a significant amount of research has been undertaken from different disciplinary angles to understand this challenge-from computer science to law to sociology-none of these fields have been able to resolve the problem on their own terms. For instance, computational methods to verify and certify bias-free data

View PDFchevron_right
Algorithmic Fairness
Dana Pessach

2020

An increasing number of decisions regarding the daily lives of human beings are being controlled by artificial intelligence (AI) algorithms in spheres ranging from healthcare, transportation, and education to college admissions, recruitment, provision of loans and many more realms. Since they now touch on many aspects of our lives, it is crucial to develop AI algorithms that are not only accurate but also objective and fair. Recent studies have shown that algorithmic decision-making may be inherently prone to unfairness, even when there is no intention for it. This paper presents an overview of the main concepts of identifying, measuring and improving algorithmic fairness when using AI algorithms. The paper begins by discussing the causes of algorithmic bias and unfairness and the common definitions and measures for fairness. Fairness-enhancing mechanisms are then reviewed and divided into pre-process, in-process and post-process mechanisms. A comprehensive comparison of the mechani...

View PDFchevron_right
Fair Enough? A map of the current limitations of the requirements to have "fair" algorithms
Ilaria Penco

arXiv (Cornell University), 2023

View PDFchevron_right
Fairness in Algorithmic Decision-making
Adam Harkens

Amicus Curiae, 2019

This article discusses conceptions of fairness in algorithmic decision-making, within the context of the UK’s legal system. Using practical operational examples of algorithmic tools, itargues that such practices involve inherent technical trade-offs over multiple, competing notions of fairness, which are further exacerbated by policy choices made by those public authorities who use them. This raises major concerns regarding the ability of such choices to affect legal issues in decision-making, and transform legal protections, without adequate legal oversight, or a clear legal framework. This is not to say that the law does not have the capacity to regulate and ensure fairness, but that a more expansive idea of its function is required.

View PDFchevron_right

References (53)

  1. Aas, K., Jullum, M., & Løland, A. (2021). Explaining individual predictions when features are depend- ent: More accurate approximations to Shapley values. Artificial Intelligence, 298, 103502. https:// doi. org/ 10. 1016/j. artint. 2021. 103502
  2. ACLU California Action. (2020). AB 256. ACLU California Action. https:// acluc alact ion. org/ bill/ ab-256/ Abdollahi, B., & Nasraoui, O. (2018). Transparency in fair machine learning: The case of explain- able recommender systems. In J. Zhou & F. Chen (Eds.), Human and machine learning: Vis- ible, explainable, trustworthy and transparent (pp. 21-35). Springer. https:// doi. org/ 10. 1007/ 978-3-319-90403-0_2
  3. Adadi, A., & Berrada, M. (2018). Peeking inside the black-box: A survey on explainable artificial intel- ligence (XAI). IEEE Access, 6, 52138-52160. https:// doi. org/ 10. 1109/ ACCESS. 2018. 28700 52
  4. Agyeman, J. (2021, March 9). How urban planning and housing policy helped create 'food apartheid' in US cities. The Conversation. http:// theco nvers ation. com/ how-urban-plann ing-and-housi ng-policy- helped-create-food-apart heid-in-us-cities-154433
  5. Aivodji, U., Arai, H., Fortineau, O., Gambs, S., Hara, S., & Tapp, A. (2019). Fairwashing: The risk of rationalization. In Proceedings of the 36th international conference on machine learning, 2019 (pp. 161-170). https:// proce edings. mlr. press/ v97/ aivod ji19a. html
  6. Ananny, M., & Crawford, K. (2018). Seeing without knowing: Limitations of the transparency ideal and its application to algorithmic accountability. New Media and Society, 20(3), 973-989. https:// doi. org/ 10. 1177/ 14614 44816 676645
  7. Barabas, C., Dinakar, K., Ito, J., Virza, M., & Zittrain, J. (2018). Interventions over predictions: Refram- ing the ethical debate for actuarial risk assessment. arXiv: 1712. 08238 [Cs, Stat]. http:// arxiv. org/ abs/ 1712. 08238
  8. Barocas, S., & Selbst, A. D. (2016). Big data's disparate impact. SSRN Electronic Journal. https:// doi. org/ 10. 2139/ ssrn. 24778 99
  9. Barocas, S., Selbst, A. D., & Raghavan, M. (2020). The hidden assumptions behind counterfactual expla- nations and principal reasons. In Proceedings of the 2020 conference on fairness, accountability, and transparency, 2020 (pp. 80-89). https:// doi. org/ 10. 1145/ 33510 95. 33728 30
  10. Begley, T., Schwedes, T., Frye, C., & Feige, I. (2020). Explainability for fair machine learning. arXiv: 2010. 07389 [Cs, Stat]. http:// arxiv. org/ abs/ 2010. 07389
  11. Christman, J. (2020). Autonomy in Moral and Political Philosophy. In E. N. Zalta (Ed.), The Stanford Encyclopedia of Philosophy (Fall 2020). Metaphysics Research Lab, Stanford University. https:// plato. stanf ord. edu/ archi ves/ fall2 020/ entri es/ auton omy-moral/
  12. Citron, D. K., & Pasquale, F. A. (2014). The scored society: Due process for automated predictions (SSRN Scholarly Paper ID 2376209). Social Science Research Network. https:// papers. ssrn. com/ abstr act= 23762 09
  13. Datta, A., Sen, S., & Zick, Y. (2016). Algorithmic transparency via quantitative input influence: Theory and experiments with learning systems. In 2016 IEEE symposium on security and privacy (SP), 2016 (pp. 598-617). https:// doi. org/ 10. 1109/ SP. 2016. 42
  14. Dodge, J., Liao, Q. V., Zhang, Y., Bellamy, R. K. E., & Dugan, C. (2019). Explaining models: An empirical study of how explanations impact fairness judgment. In Proceedings of the 24th international confer- ence on intelligent user interfaces, 2019 (pp. 275-285). https:// doi. org/ 10. 1145/ 33012 75. 33023 10
  15. Doshi-Velez, F., & Kim, B. (2017). Towards a rigorous science of interpretable machine learning. arXiv: 1702. 08608 [Cs, Stat]. http:// arxiv. org/ abs/ 1702. 08608
  16. Frye, C., Rowat, C., & Feige, I. (2020). Asymmetric Shapley values: Incorporating causal knowledge into model-agnostic explainability. In Advances in neural information processing systems, 2020 (Vol. 33, pp. 1229-1239). https:// proce edings. neuri ps. cc/ paper/ 2020/ hash/ 0d770 c496a a3da6 d2c3f 2bd19 e7b9d 6b-Abstr act. html
  17. Galhotra, S., Pradhan, R., & Salimi, B. (2021). Explaining black-box algorithms using probabilistic contras- tive counterfactuals (arXiv: 2103. 11972). https:// doi. org/ 10. 48550/ arXiv. 2103. 11972
  18. Greiner, D. J. (2008). Casual inference in civil rights litigation. Harvard Law Review, 122, 533.
  19. Grgić-Hlača, N., Zafar, M. B., Gummadi, K. P., & Weller, A. (2018). Beyond distributive fairness in algorith- mic decision making: feature selection for procedurally fair learning. In Proceedings of the AAAI con- ference on artificial intelligence, 2018 (Vol. 32(1), Article 1). https:// doi. org/ 10. 1609/ aaai. v32i1. 11296
  20. Hacker, P. (2022). The European AI Liability Directives -Critique of a Half-Hearted Approach and Lessons for the Future. SSRN Electronic Journal. https:// doi. org/ 10. 2139/ ssrn. 42797 96
  21. Haibe-Kains, B., Adam, G. A., Hosny, A., Khodakarami, F., Waldron, L., Wang, B., McIntosh, C., Gold- enberg, A., Kundaje, A., Greene, C. S., Broderick, T., Hoffman, M. M., Leek, J. T., Korthauer, K., Huber, W., Brazma, A., Pineau, J., Tibshirani, R., Hastie, T.,…,Aerts, H. J. W. L. (2020). Transpar- ency and reproducibility in artificial intelligence. Nature, 586(7829), 7829. https:// doi. org/ 10. 1038/ s41586-020-2766-y
  22. Heskes, T., Sijben, E., Bucur, I. G., & Claassen, T. (2020). Causal Shapley values: Exploiting causal knowl- edge to explain individual predictions of complex models (arXiv: 2011. 01625). https:// doi. org/ 10. 48550/ arXiv. 2011. 01625
  23. Hill, R. K. (2016). Genealogy. In Routledge encyclopedia of philosophy (1st ed.). Routledge. https:// doi. org/ 10. 4324/ 97804 15249 126-DE024-1
  24. Hu, L. (2019). Disparate causes, Pt. I. Phenomenal World. https:// www. pheno menal world. org/ analy sis/ dispa rate-causes-i/
  25. Jung, Y., Kasiviswanathan, S., Tian, J., Janzing, D., Bloebaum, P., & Bareinboim, E. (2022). On measur- ing causal contributions via do-interventions. In Proceedings of the 39th international conference on machine learning, 2022 (pp. 10476-10501). https:// proce edings. mlr. press/ v162/ jung2 2a. html
  26. Karimi, A.-H., Barthe, G., Schölkopf, B., & Valera, I. (2021). A survey of algorithmic recourse: Definitions, formulations, solutions, and prospects (arXiv: 2010. 04050). arXiv. http:// arxiv. org/ abs/ 2010. 04050
  27. Karimi, A.-H., Schölkopf, B., & Valera, I. (2020). Algorithmic recourse: From counterfactual explanations to interventions (arXiv: 2002. 06278). https:// doi. org/ 10. 48550/ arXiv. 2002. 06278
  28. Kohler-Hausmann, I. (2019). Eddie Murphy and the dangers of counterfactual causal thinking about detect- ing racial discrimination. Northwestern University Law Review, 113(5), 1163-1227.
  29. Leben, D. (2023). Explainable AI as evidence of fair decisions. Frontiers in Psychology. https:// doi. org/ 10. 3389/ fpsyg. 2023. 10694 26
  30. Lundberg, S. (2018). Explaining quantitative measures of fairness-SHAP latest documentation. https:// shap. readt hedocs. io/ en/ latest/ examp le_ noteb ooks/ overv iews/ Expla ining% 20qua ntita tive% 20mea sures% 20of% 20fai rness. html
  31. Lundberg, S. M., & Lee, S.-I. (2017). A unified approach to interpreting model predictions. In Advances in neural information processing systems, 2017 (Vol. 30). https:// papers. nips. cc/ paper/ 2017/ hash/ 8a20a 86219 78632 d76c4 3dfd2 8b677 67-Abstr act. html
  32. Mitchell, S., Potash, E., Barocas, S., D'Amour, A., & Lum, K. (2021). Prediction-based decisions and fair- ness: A catalogue of choices, assumptions, and definitions. Annual Review of Statistics and Its Applica- tion, 8(1), annurev-statistics-042720-125902. https:// doi. org/ 10. 1146/ annur ev-stati stics-042720-125902
  33. Mokander, J., & Floridi, L. (2021). Ethics-based auditing to develop trustworthy AI. SSRN Scholarly Paper ID 3788841. Social Science Research Network. https:// papers. ssrn. com/ abstr act= 37888 41
  34. Mökander, J., Morley, J., Taddeo, M., & Floridi, L. (2021). Ethics-based auditing of automated decision- making systems: Nature, scope, and limitations. Science and Engineering Ethics, 27(4), 44. https:// doi. org/ 10. 1007/ s11948-021-00319-4
  35. Nabi, R., & Shpitser, I. (2018). Fair inference on outcomes (arXiv: 1705. 10378). http:// arxiv. org/ abs/ 1705. 10378
  36. Nannini, L., Balayn, A., & Smith, A. L. (2023). Explainability in AI policies: A critical review of communi- cations, reports, regulations, and standards in the EU, US, and UK (arXiv: 2304. 11218). https:// doi. org/ 10. 48550/ arXiv. 2304. 11218
  37. Pearl, J. (2000). Causality: Models, reasoning, and inference. Cambridge University Press.
  38. Pearl, J. (2009). Causality (2nd ed.). Cambridge University Press. https:// doi. org/ 10. 1017/ CBO97 80511 803161
  39. Perrino, J. (2020, July 2). "Redlining" and health indicators: Decisions made 80 years ago have health con- sequences today. NCRC. https:// ncrc. org/ redli ning-and-health-indic ators-decis ions-made-80-years-ago- have-health-conse quenc es-today/
  40. Prince, A. E. R., & Schwarcz, D. (2019). Proxy discrimination in the age of artificial intelligence and big data. Iowa Law Review, 105, 1257.
  41. Roberts, H., Ziosi, M., Osborne, C., Saouma, L., Belias, A., Buchser, M., Casovan, A., Kerry, C., Meltzer, J., Mohit, S., Ouimette, M.-E., Renda, A., Stix, C., Teather, E., Woodhouse, R., & Zeng, Y. (2023). A com- parative framework for AI regulatory policy. https:// ceimia. org/ wp-conte nt/ uploa ds/ 2023/ 02/ Compa rative-Frame work-for-AI-Regul atory-Policy. pdf
  42. Rueda, J., Delgado, J., Parra Jounou, I., Hortal Carmona, J., Ausín, T., & Rodríguez-Arias, D. (2022). "Just" accuracy? Procedural fairness demands explainability in AI-based medical resource allocations. AI and Society. https:// doi. org/ 10. 1007/ s00146-022-01614-9
  43. Shapley, L. S. (1951). A value for N-person games. RAND Corporation. https:// www. rand. org/ pubs/ papers/ P295. html
  44. Solon, B. (Director). (2022, August 19). SRA22 Day 3-Keynote talk with Solon Barocas. https:// www. youtu be. com/ watch?v= Ft5rK 1tTYyw
  45. Štrumbelj, E., & Kononenko, I. (2014). Explaining prediction models and individual predictions with fea- ture contributions. Knowledge and Information Systems, 41(3), 647-665. https:// doi. org/ 10. 1007/ s10115-013-0679-x
  46. Sundararajan, M., & Najmi, A. (2020). The many Shapley values for model explanation. In Proceedings of the 37th international conference on machine learning, 2020 (pp. 9269-9278). https:// proce edings. mlr. press/ v119/ sunda raraj an20b. html
  47. Venkatasubramanian, S., & Alfano, M. (2020). The philosophical basis of algorithmic recourse. In Proceed- ings of the 2020 conference on fairness, accountability, and transparency, 2020 (pp. 284-293). https:// doi. org/ 10. 1145/ 33510 95. 33728 76
  48. Verma, S., Dickerson, J., & Hines, K. (2020). Counterfactual explanations for machine learning: A review. arXiv: 2010. 10596 [Cs, Stat]. http:// arxiv. org/ abs/ 2010. 10596
  49. Wachter, S., Mittelstadt, B., & Russell, C. (2017). Counterfactual explanations without opening the black box: Automated decisions and the GDPR. Harvard Journal of Law and Technology, 31(2), 841-888.
  50. Wachter, S., Mittelstadt, B., & Russell, C. (2021). Bias preservation in machine learning: The legality of fair- ness metrics under EU Non-Discrimination Law (SSRN Scholarly Paper ID 3792772). Social Science Research Network. https:// doi. org/ 10. 2139/ ssrn. 37927 72
  51. Wallin, D. E. (1992). Legal recourse and the demand for auditing. The Accounting Review, 67(1), 121-147.
  52. Wang, J., Wiens, J., & Lundberg, S. (2021). Shapley flow: A graph-based approach to interpreting model predictions (arXiv: 2010. 14592). https:// doi. org/ 10. 48550/ arXiv. 2010. 14592
  53. Zhou, J., Chen, F., & Holzinger, A. (2022). Towards explainability for AI fairness. In A. Holzinger, R. Goe- bel, R. Fong, T. Moon, K.-R. Müller & W. Samek (Eds.), xxAI-Beyond explainable AI: International workshop, held in conjunction with ICML 2020: Revised and extended papers, July 18, 2020, Vienna, Austria (pp. 375-386). Springer. https:// doi. org/ 10. 1007/ 978-3-031-04083-2_ 18

Related papers

An Epistemic Lens on Algorithmic Fairness
Elizabeth Edenberg

Equity and Access in Algorithms, Mechanisms, and Optimization

In this position paper, we introduce a new epistemic lens for analyzing algorithmic harm. We argue that the epistemic lens we propose herein has two key contributions to help reframe and address some of the assumptions underlying inquiries into algorithmic fairness. First, we argue that using the framework of epistemic injustice helps to identify the root causes of harms currently framed as instances of representational harm. We suggest that the epistemic lens offers a theoretical foundation for expanding approaches to algorithmic fairness in order to address a wider range of harms not recognized by existing technical or legal definitions. Second, we argue that the epistemic lens helps to identify the epistemic goals of inquiries into algorithmic fairness. There are two distinct contexts within which we examine algorithmic harm: at times, we seek to understand and describe the world as it is, and, at other times, we seek to build a more just future. The epistemic lens can serve to direct our attention to the epistemic frameworks that shape our interpretations of the world as it is and the ways we envision possible futures. Clarity with respect to which epistemic context is relevant in a given inquiry can further help inform choices among the different ways of measuring and addressing algorithmic harms. We introduce this framework with the goal of initiating new research directions bridging philosophical, legal, and technical approaches to understanding and mitigating algorithmic harms. CCS CONCEPTS • Computing methodologies → Philosophical/theoretical foundations of artificial intelligence.

View PDFchevron_right
Fairness in Algorithmic Decision Making: An Excursion Through the Lens of Causality
Aria Khademi, Vasant G Honavar

2019

As virtually all aspects of our lives are increasingly impacted by algorithmic decision making systems, it is incumbent upon us as a society to ensure such systems do not become instruments of unfair discrimination on the basis of gender, race, ethnicity, religion, etc. We consider the problem of determining whether the decisions made by such systems are discriminatory, through the lens of causal models. We introduce two definitions of group fairness grounded in causality: fair on average causal effect (FACE), and fair on average causal effect on the treated (FACT). We use the Rubin-Neyman potential outcomes framework for the analysis of cause-effect relationships to robustly estimate FACE and FACT. We demonstrate the effectiveness of our proposed approach on synthetic data. Our analyses of two real-world data sets, the Adult income data set from the UCI repository (with gender as the protected attribute), and the NYC Stop and Frisk data set (with race as the protected attribute), show that the evidence of discrimination obtained by FACE and FACT, or lack thereof, is often in agreement with the findings from other studies. We further show that FACT, being somewhat more nuanced compared to FACE, can yield findings of discrimination that differ from those obtained using FACE.

View PDFchevron_right
An Empirical Analysis of Racial Categories in the Algorithmic Fairness Literature
Amina Abdu

2023 ACM Conference on Fairness, Accountability, and Transparency

Recent work in algorithmic fairness has highlighted the challenge of defining racial categories for the purposes of anti-discrimination. These challenges are not new but have previously fallen to the state, which enacts race through government statistics, policies, and evidentiary standards in anti-discrimination law. Drawing on the history of state race-making, we examine how longstanding questions about the nature of race and discrimination appear within the algorithmic fairness literature. Through a content analysis of 60 papers published at FAccT between 2018 and 2020, we analyze how race is conceptualized and formalized in algorithmic fairness frameworks. We note that differing notions of race are adopted inconsistently, at times even within a single analysis. We also explore the institutional influences and values associated with these choices. While we find that categories used in algorithmic fairness work often echo legal frameworks, we demonstrate that values from academic computer science play an equally important role in the construction of racial categories. Finally, we examine the reasoning behind different operationalizations of race, finding that few papers explicitly describe their choices and even fewer justify them. We argue that the construction of racial categories is a value-laden process with significant social and political consequences for the project of algorithmic fairness. The widespread lack of justification around the operationalization of race reflects institutional norms that allow these political decisions to remain obscured within the backstage of knowledge production.

View PDFchevron_right
How AI developers can assure algorithmic fairness
Hossana Twinomurinzi

Discover Artificial Intelligence, 2023

Artificial intelligence (AI) has rapidly become one of the technologies used for competitive advantage. However, there are also growing concerns about bias in AI models as AI developers risk introducing bias both unintentionally and intentionally. This study, using a qualitative approach, investigated how AI developers can contribute to the development of fair AI models. The key findings reveal that the risk of bias is mainly because of the lack of gender and social diversity in AI development teams, and haste from AI managers to deliver much-anticipated results. The integrity of AI developers is also critical as they may conceal bias from management and other AI stakeholders. The testing phase before model deployment risks bias because it is rarely representative of the diverse societal groups that may be affected. The study makes practical recommendations in four main areas: governance, social, technical, and training and development processes. Responsible organisations need to take deliberate actions to ensure that their AI developers adhere to fair processes when developing AI; AI developers must prioritise ethical considerations and consider the impact their models may have on society; partnerships between AI developers, AI stakeholders, and society that might be impacted by AI models should be established; and AI developers need to prioritise transparency and explainability in their models while ensuring adequate testing for bias and corrective measures before deployment. Emotional intelligence training should also be provided to the AI developers to help them engage in productive conversations with individuals outside the development team.

View PDFchevron_right
Algorithmic Fairness in AI
Suzana Alpsancar

Business & Information Systems Engineering, 2023

View PDFchevron_right

Related topics

  • Artificial Intelligence
  • Ethics
  • Epistemology
  • Machine Learning
  • XAI - Explainable Artificial Int...
    • 580 California St., Suite 400
    San Francisco, CA, 94104
    © 2025 Academia. All rights reserved