IEEE Technology and Society Magazine - March 2021 - 61

situations of emergency, but public agencies need
to provide the required information or protocols to
protect such data, as well as explaining how this will
be used for the common good.
Currently, there is not a mature, global-standards
body to help shape global governance of AI. Given
that public sector organizations are aligned on the
common good principle, these entities are best
placed to come up with standards of ethics for AI
that are beneficial for all. At the same time, no single organization or institution is capable of shaping
the governance guidelines for AI, given that we also
have to consider ethics, legal, socioeconomic, and
cultural aspects are sensitive to local variations. This
is why the role of a global consortium, comprising
multiple government entities, will be essential to provide a global reference for AI ethics.
There are currently proposals to address this issue
with methodologies based on risk impact assessments.
In Europe, Article 35 of the General Data Protection
Regulation [53] defines a protocol to implement data
protection impact assessments for those technologies
that process personal data. Similar initiatives can be
found now to audit the impact of algorithms based on
automated decision-making, the so-called Algorithm
Impact Assessment (e.g., [54] and [55]), which are
expected to become one of the main facilitators to
encourage the commercialization of transparent and
accountable of AI systems [56].

fAIr LAC Jalisco
As we explained in the " AI digital transformation
in public services " and " Ethics and AI " sections, AI
offers great development potential for society and
for improving the lives of citizens, but not without
the risks associated with the misuse or poor design of
these technologies or the biased data they are feed
with. With these challenges in mind, to promote the
responsible and ethical use of AI, improve social services, and mitigate the growing social inequality in
the region, the fAIr LAC project in Latin America and
the Caribbean was born.

fAIr Jalisco
A partnership among the IADB, the State Government of Jalisco and Instituto Tecnologico y de Estudios Superiores de Monterrey (ITESM) are piloting
a comprehensive AI center in Jalisco, Mexico, to
accelerate the responsible use of AI for social policies and impact entrepreneurship.

March 2021

The first pilot is centered in an AI-based application for the health care public sector. It is supporting an inclusive approach of AI toward vulnerable
and (possibly) excluded populations. Health care is
clearly one of the most dynamic, yet challenging sectors in Mexico and the LAC region. In Mexico, health
care is a universal service and is free.
Nevertheless, in the whole region, health care
faces major challenges of quality, accessibility, and
affordability for a large section of the population:
* Shortage of qualified health care professionals
and services like qualified doctors, nurses, technicians, and infrastructure.
* Nonuniform accessibility to health care services
across the region with physical access and connectivity continuing to be the major barrier to
both preventive and curative health services, and
the glaring disparity between rural and urban
areas in all countries.
AI-based solutions can augment the scarce personnel's capacities to give services and extend the
power and the efficiency of the lab facilities; help
overcome the barriers to access and solve the accessibility problem through early detection, diagnostic,
decision-making, and treatment, cater to a large part
of the Jalisco region. Those tools are not meant to
substitute nor to eliminate human experts but augment their capacities. Still, locally, there is little
experience in organizing the data acquisition, the
system validation and training of the involved clinicians. It is also necessary to build trust in the system
performance.

Use case: AI for diabetic retinopathy
Diabetic retinopathy (DR) is a complication of
diabetes mellitus (DM) [57] and is the primary cause
of blindness in developing and developed countries
[58]. According to [57], early-stage DR detection
reduces the risk of vision loss and blindness in DM
patients. In Mexico, DR is especially challenging due
to the following:
* High prevalence of diabetes: The World Health
Organization (WHO) reported that the prevalence of diabetes in Mexico is around 10.4% in
2016. Mexico also has one of the highest rates of
deaths attributable to diabetes [59]. According
to INEGI, there were 101,257 death cases caused
by diabetes in Mexico in 2018.

https://www.who.int/diabetes/country-profiles/mex_en.pdf https://www.inegi.org.mx/sistemas/olap/registros/vitales/mortalidad/tabulados/ConsultaMortalidad.asp

IEEE Technology and Society Magazine - March 2021

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