Distributive Justice, Institutionalism, and Autonomous Vehicles

Abstract

AbstractMuch of the philosophical discussion concerning the permissibility or regulation of autonomous vehicles has focused on interpersonal ethics, including and especially the question of whether we can program a vehicle to behave as an ideal moral agent would when faced with a dilemma where a fatal crash is unavoidable. However, there are normative issues surrounding autonomous vehicles that go beyond the single, isolated decision of how to act when faced with this kind of “trolley problem.” One should also ask how different autonomous vehicle designs and regulatory regimes support or undermine social and distributive justice. In this paper, the author describes three different types of justice-based worries. First, there are justice-based considerations that are internal to autonomous vehicle deployment where we determine whether autonomous vehicles will behave in ways that distribute risks fairly among relevant classes of people. That is, it would be problematic if we purchased superior trolley problem performance at the cost of dramatically increased risks in other contexts. Second, determining whether and to what extent autonomous vehicle deployment scenarios generate conflicts with other domains of justice is necessary. Finally, there are justice-based worries concerned with dual-use technologies. If superior performance on trolley problems requires the development of technologies that can be used in less benign contexts, then those risks ought to be mitigated. After describing these three sources of concern, the author then offers some potential principles—using concepts borrowed from Rawlsian distributive justice—that might help guide us through these conflicts and trade-offs.