IEEE Technology and Society Magazine - March 2021 - 66

AI for a Fair, Just, and Equitable World

to be well in line with local ethical, legal, and cultural values, uses, and customs. Respect for the
individuals' dignity ought to be a priority.
* Justifiable: Designers have to prioritize the transparency of how to design and implement the
model. Besides, special care has to be put in the
mechanisms for justification and interpretability
of its decisions and behavior. In our case, the
explanation tool is still to be developed.
* Fair and nondiscriminatory: Designers have to
consider and definitively avoid its potential to
have discriminatory effects on individuals and
social groups-this a paramount issue in Latin
America-mitigate biases which may influence
the model's outcome. Also, be aware of fairness
issues throughout the design and implementation life cycle. Fairness values have to be agreed
not only with the administration, but also with
the citizens and their organizations. Explanations
about the outcomes ought to be clear and accessible for the general public.
* Worthy of public trust: The designers have to
guarantee in as much as possible the safety,
accuracy, reliability, security, and robustness of
its product. In particular, data security and reliability are crucial elements in building trust. We
propose experimentation through sandboxes,
where citizens and NGOs can actively participate
and be empowered to inform other citizens and
give feedback.
As AI scientists we are in the midst of a discussion to decide how algorithms can, and ought, to
be transparent to then be governable-including
algorithms based on automated decision-making,
such as those based in ML and AI. Governmental
agencies, like the US Congress, are already working
in this direction [5].
For many systemic reasons, governments have
much room from improvement when it comes to
technological advancement, and AI-based tools
alone will not solve those problems. But, AI is a
game-changing technology. It is time to use it for the
benefit of the citizen. Implementation and use of a
human-centric and trustworthy AI in citizen services
may also become an indicator of how the public
administration can leverage other emerging digital
tools. The development of an ethical AI is a question

of urgency.

Acknowledgments
The work of Prof. Ulises Cortés and Dr. Atia Cortés
was supported in part by the project A European
AI On Demand Platform and Ecosystem (AI4EU)
H2020-ICT-26 # under Grant 825619. The views
expressed in this article are not necessarily those
of the consortium AI4EU. Prof. Cortes (SNI-III) and
Dr. E. Ulises Moya-Sánchez (SNI-I) are members of
Sistema Nacional de Investigadores (CONACyT,
Mexico).

 References
[1] Data Quality and Artificial Intelligence-Mitigating
Bias and Error to Protect Fundamental Rights, Eur.
Union Agency Fundam. Rights, Vienna, Austria, 2019.
[2] Commissioner for Human Rights Council of Europe.
(2020). Unboxing Artificial Intelligence: 10 Steps to
Protect Human Rights. [Online]. Available: https://www.
coe.int/en/web/commissioner/-/unboxing-artificialintelligence-10-steps-to-protect-human-rights
[3] D. Leslie. (2019). Understanding Artificial Intelligence
Ethics and Safety: A Guide for the Responsible
Design and Implementation of AI Systems in the
Public Sector. U.K. [Online]. Available: https://doi.
org/10.5281/zenodo.3240529
[4] U S E O of the President and J. Podesta, " Big data:
Seizing opportunities, preserving values, " White
House, Executive Office President, Washington, DC,
USA, Tech. Rep., 2014.
[5] US Congress. (2019). Algorithmic Accountability Act
of 2019. [Online]. Available: https://www.congress.gov/
bill/116th-congress/ house-bill/2231/all-info
[6] J. M. Lassalle, Ciberleviatan. El colapso de la
democracia liberal frente a la revolución Digital, vol.
73. Barcelona, Spain: ARPA, 2019.
[7] R. Bellaby, " Going dark: Anonymising technology in
cyberspace, " Ethics Inf. Technol., vol. 20, pp. 189-204,
Jun. 2018, doi: 10.1007/s10676-018-9458-4.
[8] S. Wachter, B. Mittelstadt, and L. Floridi, " Why a right
to explanation of automated decision-making does
not exist in the general data protection regulation, " Int.
Data Privacy Law, vol. 7, no. 2, pp. 76-99, May 2017.
[9] (2019). fAIr LAC. [Online]. Available: https://www.iadb.
org/en/fairlac
[10] M. Cabrol et al., " fAIr LAC: Responsible and
widespread adoption of artificial intelligence in Latin
America and the Caribbean, " in Proc. IABD, 2020, doi:
10.18235/0002169.
[11] G. Viale Pereira et al., " Increasing collaboration and
participation in smart city governance: A cross-case

IEEE Technology and Society Magazine

https://www.coe.int/en/web/commissioner/-/unboxing-artificial-intelligence-10-steps-to-protect-human-rights https://www.coe.int/en/web/commissioner/-/unboxing-artificial-intelligence-10-steps-to-protect-human-rights https://www.coe.int/en/web/commissioner/-/unboxing-artificial-intelligence-10-steps-to-protect-human-rights https://doi.org/10.5281/zenodo.3240529 https://doi.org/10.5281/zenodo.3240529 https://www.congress.gov/bill/116th-congress/house-bill/2231/all-info https://www.congress.gov/bill/116th-congress/house-bill/2231/all-info https://www.iadb.org/en/fairlac https://www.iadb.org/en/fairlac

IEEE Technology and Society Magazine - March 2021

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