The problem of ethical decision making, which has long been a grand challenge for AI [23], has recently caught the public imagination. Perhaps its best-known manifestation is a modern variant of the classic trolley problem [10]: An autonomous vehicle has a brake failure, leading to an accident with inevitably tragic consequences; due to the vehicle's superior perception and computation capabilities, it can make an informed decision. Should it stay its course and hit a wall, killing its three passengers, one of whom is a young girl? Or swerve and kill a male athlete and his dog, who are crossing the street on a red light? A notable paper by Bonnefon et al. [2] has shed some light on how people address such questions, and even former US President Barack Obama has weighed in.

When a semi-autonomous car crashes and harms someone, how are blame and causal responsibility distributed across the human and machine drivers? In this article, we consider cases in which a pedestrian was hit and killed by a car being operated under shared control of a primary and a secondary driver. We find that when only one driver makes an error, that driver receives the blame and is considered causally responsible for the harm, regardless of whether that driver is a machine or a human. However, when both drivers make errors in cases of shared control between a human and a machine, the blame and responsibility attributed to the machine is reduced. This finding portends a public under-reaction to the malfunctioning AI components of semi-autonomous cars and therefore has a direct policy implication: a bottom-up regulatory scheme (which operates through tort law that is adjudicated through the jury system) could fail to properly regulate the safety of shared-control vehicles; instead, a top-down scheme (enacted through federal laws) may be called for.

We present a general approach to automating ethical decisions, drawing on machine learning and computational social choice. In a nutshell, we propose to learn a model of societal preferences, and, when faced with a specific ethical dilemma at runtime, efficiently aggregate those preferences to identify a desirable choice. We provide a concrete algorithm that instantiates our approach; some of its crucial steps are informed by a new theory of swap-dominance efficient voting rules. Finally, we implement and evaluate a system for ethical decision making in the autonomous vehicle domain, using preference data collected from 1.3 million people through the Moral Machine website.