Paper Type
ERF
Abstract
Artificial intelligence (AI) is transforming the way humans engage in intimate relationships, taking on roles such as friends, romantic partners, pets, and therapists. However, little is known about humans’ preferences between physical robots and virtual agents in these relationships. This pilot study found variations in the roles envisioned (86% of participants envisioned AI as a friend, 75% as a therapist and 72% as a mentor, 17% as a romantic or sexual partner; caregiver and pet roles stood in the middle, with greater willingness to have an AI care for an older parent or oneself than a child). Individuals strongly preferred physical robots over virtual agents in all except therapeutic and mentoring relationships. Different factors drove these preferences, such as lifelike interaction and functional utility for physical robots and accessibility and cost effectiveness for virtual agents. These findings deepen our understanding of human-AI interactions and guide the development of human-centered AI.
Paper Number
2172
Recommended Citation
Liang, Qingyu and Ollier-Malaterre, Ariane, "Physical Robot or Virtual Agent? Humans’ Preferences in Human-AI Intimate Relationships" (2025). AMCIS 2025 Proceedings. 13.
https://aisel.aisnet.org/amcis2025/sig_aiaa/sig_aiaa/13
Physical Robot or Virtual Agent? Humans’ Preferences in Human-AI Intimate Relationships
Artificial intelligence (AI) is transforming the way humans engage in intimate relationships, taking on roles such as friends, romantic partners, pets, and therapists. However, little is known about humans’ preferences between physical robots and virtual agents in these relationships. This pilot study found variations in the roles envisioned (86% of participants envisioned AI as a friend, 75% as a therapist and 72% as a mentor, 17% as a romantic or sexual partner; caregiver and pet roles stood in the middle, with greater willingness to have an AI care for an older parent or oneself than a child). Individuals strongly preferred physical robots over virtual agents in all except therapeutic and mentoring relationships. Different factors drove these preferences, such as lifelike interaction and functional utility for physical robots and accessibility and cost effectiveness for virtual agents. These findings deepen our understanding of human-AI interactions and guide the development of human-centered AI.
When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.
Comments
SIGAIAA