Ethical Challenges and Policy Approaches in Artificial Intelligence

Abstract

Artificial intelligence (AI) is transforming various fields at a rapid rate. This phenomenon has raised major ethical issues, leading to the establishment of regulations by governments, organizations, and educational institutions. In spite of the increasing rate of AI-related activities, there have been contradictions in the concepts and regulations. There have been convergences in the fundamental concepts of AI ethics, including beneficence, non-maleficence, autonomy, justice, and explicability. Empirical studies have shown that there is a good understanding of AI ethics. However, the application of the numerous mitigating techniques that have been proposed in the literature remains low. Mixed-method studies of international regulations have shown major deficiencies in the regulations, especially from the Global South and underrepresented groups. Evaluations of regulations in different sectors, including healthcare, education, and governance, have shown that there are insufficient instruments and low levels of stakeholder engagement in the execution of AI ethics. New frameworks like EMMA and Ethics by Design have shown promising strategies for the integration of ethical monitoring in the life cycle of AI development. There is a major gap between the ethics of AI and the regulations that can practically be implemented.