Addressing Algorithmic Bias
Abhishek Benedict Kumar, Karun Sanjaya · Advances in computational intelligence and robotics book series · 2025
As artificial intelligence (AI) systems increasingly make decisions in areas such as justice, health, and finance, issues related to algorithmic biases have risen to prominence when discussing equity and inclusion. This chapter investigates how prejudices are built into data and model specifications, and how they can contribute to further entrenching social disparities. From a legal and ethical perspective, it examines the shortcoming that current anti-discrimination laws face when confronted with AI-generated harm. The chapter provides examples of actual cases, such as COMPAS risk assessment tool, to illustrate the real-life implications of biased algorithms. It calls for holistic approaches that include legal remedies, algorithmic transparency, participatory governance, and anti-discrimination data practices. By integrating the development of AI into fair, accountable, and human rights respecting principles, the chapter emphasize the immediate necessity of ensuring that technological development is aligned with just and inclusive development.