지각된 신뢰성의 역할: 인간과 AI 에이전트 선택에서 서비스 영역의 상호작용 효과

With the rapid advancement of artificial intelligence (AI) technology, the adoption of AI agents has been expanding across various service domains. However, understanding of when and why users prefer human or AI agents remains limited. This study aimed to examine the moderating effect of service domain on user preferences between human and AI agents and the mediating role of perceived reliability. The results showed that in healthcare services, intention to use human agents was significantly higher than toward AI agents, while in financial services, intention to use AI agents was significantly higher than toward human agents. Furthermore, human agents were perceived as more reliable in healthcare services, whereas AI agents were perceived as more reliable in financial services. Mediation analysis revealed that perceived reliability mediated the effect of the interaction between agent type and service domain on users' intention to use. These findings demonstrate that users prioritize relational reliability of humans in healthcare services where empathy and clinical experience are crucial, while prioritizing analytical reliability of AI in financial services where accuracy and objectivity are important. This confirms the significance of service domain roles in human-AI interaction. The study is expected to provide meaningful practical implications for establishing differentiated AI adoption strategies based on service characteristics and improving user perceptions of reliability.