Chapter 35: Power and inequalities: lifting the veil of ignorance in AI ethics

Abstract

This chapter argues that AI ethics needs to engage with structures of power and inequalities. It draws lessons learned from critiques AI ethics has received of ineffectiveness and highlights that a key issue is the way mainstream ethics tends to systematically neglect power relations and structures of inequalities. The chapter then argues for a situated ethics. It points to the need to implement ethics within institutional frameworks, such as research ethics processes, with mechanisms to ensure compliance. Finally, it calls for another approach to ethics for AI, an ethics that places capacity of action at its heart. As such, this chapter highlights two types of relations between ethics and power: (1) ethics as soft law where ethics seeks to limit abuses of power (power over) and (2) ethics as a constantly renewed attention and attitude promoting power of action (power to).