Robots, Chatbots, Self-Driving Cars: Perceptions of Mind and Morality Across Artificial Intelligences

TL;DR

This preregistered online study assesses perceived mind (agency and experience) and morality (moral agency and moral patiency) across 14 AIs and 12 non-AIs with 975 US participants. It finds AIs generally show low-to-moderate agency and very low experience, while moral attributions—particularly moral agency—are higher and vary substantially across AIs (e.g., Tesla FSD highest in moral agency; Jennie high in patiency). Differences exist both within AIs and between AI and non-AI entities, with mind and morality related yet not identical; familiarity with AI also modulates moral patiency. The authors discuss design implications to manage perceptions and responsibility, cautioning against excessive anthropomorphism in high-stakes contexts and emphasizing alignment between perceived and actual capacities to support trust and accountability.

Abstract

AI systems have rapidly advanced, diversified, and proliferated, but our knowledge of people's perceptions of mind and morality in them is limited, despite its importance for outcomes such as whether people trust AIs and how they assign responsibility for AI-caused harms. In a preregistered online study, 975 participants rated 26 AI and non-AI entities. Overall, AIs were perceived to have low-to-moderate agency (e.g., planning, acting), between inanimate objects and ants, and low experience (e.g., sensing, feeling). For example, ChatGPT was rated only as capable of feeling pleasure and pain as a rock. The analogous moral faculties, moral agency (doing right or wrong) and moral patiency (being treated rightly or wrongly) were higher and more varied, particularly moral agency: The highest-rated AI, a Tesla Full Self-Driving car, was rated as morally responsible for harm as a chimpanzee. We discuss how design choices can help manage perceptions, particularly in high-stakes moral contexts.

Figures (6)

• Figure 1: Ratings of moral agency (capacity to do right or wrong) and moral patiency (capacity to be treated rightly or wrongly) for a subset of the entities. The AIs are shown in blue and the non-AIs are shown in red. Four AIs are shown: Roomba, rated lowest on moral agency and moral patiency; a Full Self-Driving car, rated highest on moral agency; Jennie, rated highest on moral patiency; and ChatGPT. All icons were created with DALL-E except ChatGPT which was created by Freepik and sourced from flaticon.com. Ratings of all 26 entities for both mind and morality are shown in \ref{['sec:results']}.
• Figure 2: Example images of each of the 14 artificial intelligences that were shown to participants. (a) Tesla Full Self-Driving car, (b) AgentGPT, (c) Roomba, (d) Alexa, (e) Sophia, (f) Spot, (g) Jennie, (h) Siri, (i) Cicero, (j) Replika, (k) DALL-E, (l) Wysa, (m) AlphaZero, (n) ChatGPT. Image credits and sources can be found in Appendix A.
• Figure 3: Ratings of mind (agency and experience) for the 14 AI entities.
• Figure 4: Ratings of morality (moral agency and moral patiency) for the 14 AI entities.
• Figure 5: Ratings of mind (agency and experience) for all 26 entities.