Page 89 - The Ethics of ASEAN

Ethical Challenges of our Digital World

Kelly Forbes: In ASEAN, AI ethics
are about human centricity, privacy,
fairness, explainability/transparency and
accountability

Kelly Forbes draws on her expertise in international law to
discuss the ethics of “Trustworthy AI” in ASEAN. The concept
is that in order to trust a decision made by an algorithm, we
need to know that it is lawful, ethical and robust. Trustworthy
2
AI has been adopted by various nations, regions and
intergovernmental organisations. For example, the Organisation Figure 2: Kelly Forbes,
for Economic Co-operation and Development (OECD) includes Co-Founder and Executive
Trustworthy AI as a part of its principles for investing in AI Director of Al Asia Pacific
research and development. The European Commission Institute (AIAPI)
published Ethics Guidelines for Trustworthy Artificial Intelligence
(AI), as early as 2018. 3

Ms Forbes explained the five ethical principles she uses
in advocating for Trustworthy AI in ASEAN: human-centricity,
privacy, fairness, explainability and accountability.
• Human Centricity: this refers to a human-centric approach
which places human experience at the centre of AI
technology development design and intended outcomes.
Human centricity in AI development, which focuses on its
use in our society, can foster confidence and trust while
respecting the ethics of human rights.
• Privacy: this covers the evolving issue of data protection
and individuals’ rights and authority in using one’s digital
information, profile and visibility. To be Trustworthy, the
use of AI in surveillance systems, data collection, analysis
and commercialization must ensure that personal data is
protected, and control over access and personalization of
data remains in the hands of the people.
• Fairness: in the context of AI, this refers to so-called
‘AI bias’ which can generate systematic discrimination
of specific populations based on age, gender, race and
ethnicity. Such errors reflect actual human biases in
our society, leading to the machines making wrongful
predictions and outcomes.
• Explainability & Transparency: this means that an AI
system must be able to explain clearly why it makes its
2 Trustworthy AI Report in the Asia-Pacific Region 2021 https://aiasiapacific.
org/wp-content/uploads/2021/07/2021-Trustworthy-Artificial-Intelligence-
in-the-Asia-Pacific-Region.pdf Downloaded on January 16, 2023
3 Ethics Guidelines for Trustworthy AI https://ec.europa.eu/futurium/en/ai-
alliance-consultation.1.html Retrieved on January 17, 2023

79