[Paper Review] Autonomous Vehicles Need Experimental Ethics: Are We Ready for Utilitarian Cars?

Jean‐François Bonnefon, Azim Shariff+1 more|arXiv (Cornell University)|Oct 12, 2015

Ethics and Social Impacts of AI13 references76 citations

TL;DR

This paper investigates the ethical dilemma of utilitarian decision-making in autonomous vehicles (AVs), finding that while people endorse utilitarian AVs in principle—prioritizing the greater good—most prefer non-utilitarian AVs that protect their own passengers. The study reveals a paradox: enforcing utilitarian algorithms may reduce adoption and increase casualties, undermining AV safety benefits.

ABSTRACT

Autonomous Vehicles (AVs) should reduce traffic accidents, but they will sometimes have to choose between two evils-for example, running over pedestrians or sacrificing itself and its passenger to save them. Defining the algorithms that will help AVs make these moral decisions is a formidable challenge. We found that participants to six MTurk studies approved of utilitarian AVs (that sacrifice their passengers for the greater good), and would like others to buy them, but they would themselves prefer to ride in AVs that protect their passengers at all costs. They would disapprove of enforcing utilitarian AVs, and would be less willing to buy such a regulated AV. Accordingly, regulating for utilitarian algorithms may paradoxically increase casualties by postponing the adoption of a safer technology.

Motivation & Objective

• To examine public attitudes toward utilitarian versus non-utilitarian decision-making in autonomous vehicles.
• To investigate the gap between moral approval of utilitarian AVs and personal preference for self-protecting AVs.
• To assess whether regulating for utilitarian algorithms would hinder AV adoption and increase overall casualties.
• To explore the implications of this ethical paradox for policy and technology deployment.

Proposed method

• Conducted six MTurk-based studies to survey public attitudes toward AV moral decision-making.
• Presented participants with hypothetical scenarios involving trade-offs between passenger and pedestrian safety.
• Measured approval of utilitarian AVs, willingness to buy them, and willingness to regulate for such algorithms.
• Collected data on personal preferences for self-protecting AVs versus altruistic ones.
• Analyzed the discrepancy between stated moral approval and actual behavioral preferences.
• Used survey data to model the potential real-world impact of regulating for utilitarian AVs.

Experimental results

Research questions

• RQ1 Do people approve of utilitarian AVs that sacrifice passengers to save more pedestrians?
• RQ2 Would individuals prefer to ride in AVs that prioritize their own safety over the greater good?
• RQ3 Would people support government regulation mandating utilitarian AV algorithms?
• RQ4 Would enforcing utilitarian algorithms reduce public willingness to buy AVs?
• RQ5 Could regulating for utilitarian AVs paradoxically increase overall casualties by delaying adoption?

Key findings

• Participants approved of utilitarian AVs in theory, endorsing algorithms that minimize overall harm.
• Despite this approval, participants preferred to ride in AVs that protected their own passengers at all costs.
• Participants disapproved of enforcing utilitarian algorithms through regulation.
• Regulated utilitarian AVs were less likely to be purchased, suggesting a disincentive to adoption.
• The preference gap implies that enforcing utilitarianism may delay AV deployment and increase casualties.
• The study concludes that regulatory mandates for utilitarian AVs could paradoxically reduce overall safety.