An overview of AI ethics: moral concerns through the lens of principles, lived realities and power structures

Abstract Along with the rapid development of AI systems, the literature addressing the moral concerns raised by AI—stemming from different directions like computer science, medicine, and philosophy—has substantially grown. In solely focusing on AI ethics principles and guidelines, most overviews of the field adopt a principle-based understanding of these moral concerns. However, as our review illuminates, there is more richness and diversity in the current body of literature than this dominant principle-based approach suggests. Within this vast literature, we identify three approaches by which authors tend to formulate the moral concerns raised by AI: principles, lived realities, and power structures. These approaches can be viewed as lenses through which authors investigate the field, and which each entail specific theoretical sensitivities, disciplinary traditions, and methodologies, and hence, specific strengths and weaknesses. The first “principle-based” approach takes moral concerns to be universal, stable, and fixed principles; which are globally shared, may travel between contexts and are often predetermined. What we call the “lived realities” approach foregrounds the interaction between people and AI systems, focusing on local practices and everyday experiences, generally on a micro-level. Thirdly, what we call the “power structures” approach argues for the need to account for the cultural, social, political, and economic context of AI development, hence human-AI interactions at a macro-level. In bringing together different moral frameworks, traditions, and questions, our structure may serve as a bridge for comparing AI ethics with other areas of applied ethics—considering AI systems are quickly integrated into different spheres of society.