Page 37 - AI for Good - Impact Report

AI for Good

Societal Implications

Focus area: Ethical Considerations

The ethical challenges posed by AI are complex, requiring careful consideration to ensure that
the technology is developed and deployed responsibly. Key ethical matters include bias
126
and lack of transparency or accountability. Examples of AI leading to negative social impact
include biases in the hiring process that favor men over women, more severe judgement of
127
minorities, or chatbots becoming sexist and/or racist. However, an increasing number of
128
129
advocates argue that, if implemented thoughtfully, AI has the potential to mitigate biases in
human decision-making and promote ethical considerations. 130
Governments have a crucial responsibility to address these challenges by establishing and
enforcing regulatory frameworks that set clear standards for the ethical use and deployment
of AI.

Established practice: European Union's Ethics Guidelines for Trustworthy AI
The EU’s “Ethics Guidelines for Trustworthy AI” stands out as the leading framework in
addressing the ethical challenges of AI. The guidelines emphasize that AI should “ensure
131
that the development, deployment and use of AI systems meet the seven key requirements
for Trustworthy AI: Human agency and oversight, Technical robustness and safety, Privacy and
data governance, Transparency, Diversity, non-discrimination and fairness, Environmental and
societal well-being and Accountability.”

The actionable nature of these guidelines is one of their strongest attributes. For instance,
the guidelines suggest methods for ensuring transparency in AI algorithms, such as making
decision-making processes understandable to non-experts and ensuring that AI systems can
explain their outcomes in human terms. As a major regulatory power, the EU has a significant
impact on global discussions about AI governance.
Emerging practice: UNESCO “Recommendation on the Ethics of Artificial Intelligence”

The UNESCO “Recommendation on the Ethics of Artificial Intelligence” states that “The
protection of human rights and dignity is the cornerstone of the Recommendation, based on the
advancement of fundamental principles such as transparency and fairness, always remembering
the importance of human oversight of AI systems.”
132
To support this, the text put forward a set of 10 principles: 1) Proportionality and Do no harm,
2) Safety and security, 3) Fairness and non-discrimination, 4) Sustainability, 5) Right to privacy
and Data protection, 6) Human oversight and determination, 7) Transparency and explainability,
8) Responsibility and accountability, 9) Awareness and literacy, and 10) Multi-stakeholder and
adaptive governance and collaboration. These principles are designed to establish a universal
baseline for ethical AI, ensuring that all AI systems align with fundamental human rights and
ethical standards. The UN principles are designed to be universally relevant. This broad scope
makes them accessible to countries and organizations across different cultural, economic, and
legal landscapes, including those that may not have the resources or expertise to develop their
own detailed AI ethics frameworks.

27