IEEE Technology and Society Magazine - June 2021 - 79

of particularities such as the attributes being used
in a data set, as well as the ability to compare
explanation decisions and moral rules underlying
those choices. This is, however, not solely a
technical exercise, as decision-making systems
render end-users into algorithmically constructed
data subjects. This challenge could be addressed
through a socio-technical approach that can consider
both the technical dimensions and the complex
social contexts in which these systems are
deployed. Building public confidence and greater
democratic participation in AI systems requires
ongoing development of not just explainable
AI but of better Human-AI interaction methods
and sociotechnical platforms, tools, and public
engagement to increase critical public understanding
and agency.
Discrimination-aware AI
Third, AI should not just be seen as a potential problem
causing discrimination, but also as a great opportunity
to mitigate existing issues. The fact that AI can pick
up on discrimination suggests it can be made aware of
it. For instance, AI could help spot digital forms of discrimination,
and assist in acting upon it. For this aim to
become a reality we would need, as explored in this
work, a better understanding of social, ethical, and legal
principles, as well as dialogically constructed solutions
in which this knowledge is incorporated into AI systems.
Two ways to achieve this goal are: 1) using data-driven
approaches like machine learning to actually look at previous
cases of discrimination and try to spot them in the
future and 2) using model-based and knowledge-based
AI that operationalizes the socio-ethical and legal principles
mentioned above (e.g., normative approaches that
include nondiscrimination norms as part of the knowledge
of an AI system to influence its decision making).
This would, for instance, facilitate an AI system realizing
that the knowledge it gathered or learned is resulting
in discriminatory decisions when deployed in specific
contexts. Hence, the AI system could alert an expert
human about this, and/or proactively address the issue
spotted.

Acknowledgments
This work was supported by Engineering and
Physical Sciences Research Council (EPSRC) under
Grant EP/R033188/1. It is part of the Discovering and
Attesting Digital Discrimination (DADD) project-
see https://dadd-project.org.
June 2021
 References
[1] C. O'Neil, Weapons of Math Destruction: How Big
Data increases Inequality and Threatens Democracy.
Portland, OR, USA: Broadway Books, 2017.
[2] D. Pedreshi, S. Ruggieri, and F. Turini, " Discriminationaware
data mining, " in Proc. Proc. 14th ACM SIGKDD
Int. Conf. Knowl. Discovery Data Mining (KDD), 2008,
pp. 560-568.
[3] N. Criado and J. M. Such, " Digital discrimination, " in
Algorithmic Regulation. Oxford, U.K.: OUP, 2019.
[4] D. Danks and A. London, " Algorithmic bias in autonomous
systems, " in Proc. IJCAI, 2017, pp. 4691-4697.
[5] X. Ferrer et al., " Discovering and categorising
language biases in reddit, " in Proc. Int. AAAI Conf. Web
Social Media (ICWSM), 2020.
[6] S. T. Mueller et al., " Explanation in human-AI systems:
A literature meta-review, synopsis of key ideas and
publications, and bibliography for explainable AI, " Feb.
2019, arXiv:1902.01876. [Online]. Available: http://arxiv.
org/abs/1902.01876
[7] R. Guidotti et al., " A survey of methods for explaining
black box models, " ACM Comput. Surv., vol. 51, no. 5,
pp. 93:1-93:42, Aug. 2018.
[8] R. Guidotti et al., " Factual and counterfactual
explanations for black box decision making, " IEEE
Intell. Syst., vol. 34, no. 6, pp. 14-23, Dec. 2019.
[9] S. Ruggieri, D. Pedreschi, and F. Turini, " Integrating
induction and deduction for finding evidence of
discrimination, " AI Law, vol. 18, no. 1, pp. 1-43, 2010.
[10] N. Kilbertus et al., " Avoiding discrimination through
causal reasoning, " in Proc. NIPS, 2017, pp. 656-666.
[11] R. M. Byrne, " Counterfactuals in explainable artificial
intelligence (XAI): Evidence from human reasoning, " in
Proc. IJCAI, 2019, pp. 6276-6282.
[12] T. Calders and I. Zliobaite, " Why unbiased computational
processes can lead to discriminative decision procedures, "
in Discrimination and Privacy in the Information Society.
Berlin, Germany: Springer, 2013, pp. 43-57.
[13] N. Criado, X. Ferrer, and J. M. Such, " A normative
approach to attest digital discrimination, " in Proc.
Advancing Towards SDGS Artif. Intell. Fair, Just
Equitable World Workshop 24th Eur. Conf. Artif. Intell.
(ECAI), 2020.
[14] T. Miller, " Explanation in artificial intelligence: Insights
from the social sciences, " Artif. Intell., vol. 267,
pp. 1-38, Feb. 2019.
[15] N. Grgic-Hlac et al., " Beyond distributive fairness in
algorithmic decision making, " in Proc. AAAI, 2018,
pp. 51-60. [Online]. Available: https://people.mpi-sws.
org/nghlaca/papers/fair_feature_selection.pdf
79
http://arxiv.org/abs/1902.01876 http://arxiv.org/abs/1902.01876 https://www.people.mpi-sws.org/nghlaca/papers/fair_feature_selection.pdf https://www.dadd-project.org https://www.people.mpi-sws.org/nghlaca/papers/fair_feature_selection.pdf
IEEE Technology and Society Magazine - June 2021

Table of Contents for the Digital Edition of IEEE Technology and Society Magazine - June 2021
