Satisfaction in Generative AI Supportive Interaction: The Roles of Perceived Usefulness, Perceived Supportive Interaction, and Reliance Tendency

Generative AI is increasingly used for task support, idea development, learning, problem solving, and emotional support.This study examined the associations of perceived usefulness and perceived supportive interaction with interaction satisfaction and assessed whether reliance tendency statistically accounted for part of these associations.An online survey was completed by 152 users of generative AI, and 151 respondents with complete data on the focal variables were included in the analyses.Perceived usefulness and perceived supportive interaction were each positively associated with reliance tendency and interaction satisfaction, while reliance tendency was also positively associated with interaction satisfaction.The indirect association between perceived usefulness and interaction satisfaction through reliance tendency was significant (β=.14, 95% bootstrap CI [.04, .24]),as was the indirect association between perceived supportive interaction and interaction satisfaction through reliance tendency (β=.12, 95% bootstrap CI [.04, .25]).The direct associations of both predictors with interaction satisfaction remained significant, indicating significant indirect associations alongside retained direct associations.These findings suggest that satisfaction with generative AI supportive interaction reflects both functional value and the perceived quality of conversational support, and that repeated cognitive and functional reliance may account for part of these associations.By incorporating perceived supportive interaction into a post-use satisfaction framework, this study extends technology-acceptance perspectives on generative AI use.Because the study relied on crosssectional self-report data, the indirect-association results should not be interpreted as temporal or causal processes.