생성형 AI 특성이 과업 성과와 지속 사용 행동에 미치는 영향: SOR과 TTF 이론을 중심으로

With the increasing use of Generative AI in actual task performance, this study aims to investigate how the characteristics that differentiate Generative AI from traditional AI influence user satisfaction and task-technology fit, based on the Stimulus-Organism-Response framework. Additionally, the study examines the effects of user satisfaction and task-technology fit on task performance and continued usage intention. To analyze the proposed research model, a survey was conducted with users of ChatGPT, a representative Generative AI service. A total of 315 valid responses were collected, and structural equation modeling was performed using AMOS 29.0. The results show that among the characteristics of ChatGPT (Personalization, Intelligence, Rareness, Accuracy) significantly influence both user satisfaction and task- technology fit, while Anthropomorphism has a significant effect only on user satisfaction. Furthermore, task-technology fit significantly affects user satisfaction, task performance, and continued usage intention, and user satisfaction also has a significant impact on task performance and continued usage intention. These findings contribute to expanding the theoretical scope of research on Generative AI and offer practical implications for organizations seeking to adopt generative AI to enhance task performance.