From adoption to retention: Understanding generative AI among Vietnamese students

While the initial adoption of technology is widely studied, the factors driving its long-term retention remain a critical gap. This research shifts the focus from adoption to sustained use by applying the Model for Sustained Technology Use (MSTU) to investigate generative AI engagement among Vietnamese university students. A cross-sectional survey of 100 students measured the key constructs of Habit, Satisfaction, and Perceived Usefulness, as well as their impact on Sustained Technology Use. Data were analyzed using Partial Least Squares Structural Equation Modeling (PLS-SEM). The findings reveal that Habit is the strongest direct predictor of sustained use, surpassing the influence of Perceived Usefulness (PU). While PU drives initial adoption, its effect diminishes over time, whereas Satisfaction (ST) plays a crucial mediating role in long-term engagement. These results challenge the prevailing assumption that perceived usefulness alone is sufficient to ensure long-term success. The study offers key implications for researchers and practitioners, emphasizing the importance of designing AI educational tools that seamlessly integrate into and adapt to user workflows to promote habitual use. For educators and developers, this means prioritizing features that create engaging, positive, automatic user experiences to ensure generative AI remains a retained educational resource, not a momentary novelty.