Predicting Human-Chatbot Relationships: A Mixed-Method Study on the Key Psychological Factors

2025-02-28

AI Summary

This study investigates the psychological mechanisms underlying human–social chatbot romantic relationships, focusing on romantic fantasy, loneliness, attachment style, anthropomorphism tendency, and sexual sensation seeking. Employing a mixed-methods design, it integrates a quantitative survey (N = 279), multiple regression analysis, and qualitative thematic coding of in-depth interviews with 15 participants. Results indicate that romantic fantasy is the strongest predictor of human–AI intimacy—exerting greater explanatory power than attachment style or anthropomorphism—and positively predicts perceived partner closeness in both human–AI and human–human relationships. Critically, this is the first study to demonstrate that romantic fantasy constitutes a shared core psychological mechanism across both relational domains. The findings challenge prevailing assumptions by establishing romantic fantasy not as a marginal or pathological phenomenon, but as a central, active, and constructive process in human–AI intimacy formation: all interviewees spontaneously engaged in romantic fantasy and actively pursued anthropomorphic experiences.

Technology Category

• Humans and AI: Emotional Intelligence
• Intelligent Robots: Human-Robot Interaction
• Application Domains: Humanities & Computational Social Science

Application Category

• Economics, Online Markets and Human Computation: Trust and reliance of crowd workers and data experts on GenAI
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Romantic relationships with social chatbots are becoming increasingly prevalent, raising important questions about their societal and psychological implications. Despite this growing trend, little is known about the individuals entering these synthetic relationships. This three-part study seeks to enhance understanding of the factors encompassing human-chatbot relationships by quantitatively examining the commonly discussed characteristics romantic and sexual fantasy, loneliness, attachment style, anthropomorphism, and sexual sensation seeking (Study 1A), comparing the impact of romantic and sexual fantasizing for human-chatbot versus human-human relationships (Study 1B), and providing qualitative insights into how individuals conceptualize romantic and sexual fantasies in their interactions with chatbots (Study 2). Individuals with romantic chatbot connections were interviewed (N=15) or surveyed (N=92), while participants in the comparison groups, long-distance (N=90) and cohabiting relationships (N=82), completed a questionnaire. Romantic fantasizing emerged as the strongest predictor of human-chatbot relationships, alongside anthropomorphism and anxious-avoidant attachment. Notably, romantic fantasy also predicted partner closeness across all relationship types, revealing shared psychological dynamics between human-chatbot and human-human bonds. Interviews further reinforced this, with all participants engaging in fantasy exploration while desiring their chatbot to feel as human as possible. This paper provides a novel and multifaceted examination of the psychological dynamics within human-chatbot relationships, highlighting the central yet understudied role of fantasy.

Problem

• Explores psychological factors in human-chatbot romantic relationships.
• Compares romantic and sexual fantasies in human-chatbot vs human-human relationships.
• Investigates how individuals conceptualize fantasies with chatbots through qualitative insights.

Innovation

• Quantitative analysis of psychological factors in human-chatbot relationships.
• Comparison of romantic fantasies in human-chatbot versus human-human relationships.
• Qualitative insights into romantic and sexual fantasies with chatbots.