Ethical principles in machine learning and artificial intelligence: cases from the field and possible ways forward

This paper examines the ethical implications of using machine learning (ML) and artificial intelligence (AI) in decision-making. It highlights the "black box" nature of many ML algorithms, leading to concerns about fairness, accuracy, accountability, and transparency. The paper reviews existing guidelines and literature on AI ethics, focusing on beneficence, non-maleficence, autonomy, justice, and explicability. Two case studies—risk assessment in the criminal justice system and autonomous vehicles—illustrate ethical challenges and trade-offs. The paper concludes by discussing potential ways forward, including the need for governance, interpretable models, and greater stakeholder participation.

Samuele Lo Piano