Morality on the road: Should machine drivers be more utilitarian than human drivers?

Area of Science:

• AI Ethics
• Moral Psychology
• Human-Computer Interaction

Background:

• Artificial intelligence (AI) systems, including automated vehicles (AVs), are increasingly making decisions in complex moral situations.
• These systems face moral trade-offs, necessitating an understanding of public perception regarding their decision-making processes compared to human behavior.

Purpose of the Study:

• To investigate whether people expect AI moral decision-making to differ from human moral decision-making.
• To explore the underlying reasons for any perceived differences in moral judgments between human drivers and AVs.

Main Methods:

• Six studies involving 6,805 participants examined observer judgments of human drivers and AVs in moral dilemmas.
• Dilemmas included pedestrian-only and occupant-vs-pedestrian scenarios to assess risk distribution and agent sacrifice.

Main Results:

• In pedestrian-only dilemmas, participants applied similar utilitarian norms to both human drivers and AVs.
• Participants were less accepting of AVs sacrificing their occupants than human drivers sacrificing themselves or other passengers.
• This difference was attributed to the perceived social relationship between AVs and users, not occupant agency or sacrifice voluntariness.

Conclusions:

• Public perception suggests AVs should prioritize their users' safety in moral dilemmas, diverging from expectations for human drivers.
• Understanding the social relationship's impact is crucial for developing ethically aligned AV systems.
• Findings have significant implications for AV ethics and public acceptance.