Paper Number

ICIS2025-2035

Paper Type

Complete

Abstract

Emotional bonds between users and artificial intelligence (AI) companions are increasingly common, enabled by chatbots that allow customization and communication shaped by user preferences. This study develops a feature-based taxonomy to examine how emotional engagement is fostered through specific design characteristics in AI chatbots serving companionship functions. Drawing on empirical data and interdisciplinary literature, the taxonomy identifies three meta dimensions: AI identity, interaction structure, and level of emotional engagement. Each includes distinct subdimensions and observable characteristics that influence how users perceive AI companions as emotionally present, responsive, and socially meaningful. The analysis emphasizes the role of anthropomorphic features such as adaptive language and expressive cues in enabling affective involvement. The taxonomy is applied to two use cases, Replika and Robin, illustrating how emotional connection is shaped by role framing and system design. The findings provide a conceptual foundation for future research and ethical development of emotionally responsive AI companions.

Comments

16-UserBehavior

Download

Share

COinS
 
Dec 14th, 12:00 AM

Emotional Engagement by Design: A Feature-Based Taxonomy of Chatbot Companionship

Emotional bonds between users and artificial intelligence (AI) companions are increasingly common, enabled by chatbots that allow customization and communication shaped by user preferences. This study develops a feature-based taxonomy to examine how emotional engagement is fostered through specific design characteristics in AI chatbots serving companionship functions. Drawing on empirical data and interdisciplinary literature, the taxonomy identifies three meta dimensions: AI identity, interaction structure, and level of emotional engagement. Each includes distinct subdimensions and observable characteristics that influence how users perceive AI companions as emotionally present, responsive, and socially meaningful. The analysis emphasizes the role of anthropomorphic features such as adaptive language and expressive cues in enabling affective involvement. The taxonomy is applied to two use cases, Replika and Robin, illustrating how emotional connection is shaped by role framing and system design. The findings provide a conceptual foundation for future research and ethical development of emotionally responsive AI companions.

When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.