From quality to sustainability: exploring the sustainable use of AI-powered chatbots in higher education

Abstract

With the rapid integration of smart technologies in education, artificial intelligence (AI)-powered chatbots have become increasingly popular tools for learning support. This study aims to investigate the sustainable adoption of such technologies in the higher education system by examining how chatbots' information and system quality attributes impact student perceptions and continuous use intentions. Data gathered from 230 respondents were analyzed through the partial least squares structural equation modeling (PLS-SEM) approach. The findings demonstrated that the accuracy and completeness of chatbot-generated information significantly influence perceived usefulness, while information format influences perceived ease of use. In contrast, chatbot system quality—in terms of its flexibility and timeliness—significantly influences both perceived usefulness and perceived ease of use. These favorable perceptions, along with perceived task-technology fit, significantly influence chatbot continuous-use intention. Perceived task-technology fit plays a significant mediating role in this context. Despite growing scholarly interest in chatbot integration within higher education, existing studies have predominantly employed the technology acceptance model (TAM) to explain initial adoption. This study extends the AI-assisted learning literature by integrating information quality, system quality and task-technology fit into TAM to explain students' continuous intentions toward chatbot use. It specifically examines the distinct effects of individual information quality and system quality dimensions, alongside the complementary role of task-technology fit.