Page 703 - ISC PROCEEDINGS 21.4

meaningful and responsible use of emerging technologies. Within this context, the rapid
spread of AI tools in learning is not a peripheral trend but part of a broader
transformation agenda in higher education.
In practice, however, students’ acceptance of AI tools remains uneven. Some
students view AI as a useful learning assistant capable of providing fast and flexible
support, while others remain cautious because of concerns about reliability, verification,
privacy, fairness, and the weakening of independent learning habits. Universities
therefore face a twofold challenge: they must encourage productive innovation while also
setting appropriate boundaries and support mechanisms. Identifying the factors that
shape students’ intention to use AI tools for learning is thus both academically meaningful
and practically important.
Recent scholarship has expanded quickly, particularly after the widespread diffusion
of ChatGPT and related large language models. Nevertheless, several gaps remain. First,
much of the Vietnamese evidence focuses on ChatGPT or large language models
specifically, while the broader category of AI tools for learning includes a wider range of
applications. Second, many studies explain initial acceptance but pay less attention to the
combined role of technological perceptions, social influence, learner capability, and trust
in the same model. Third, universities in Vietnam need more contextually grounded
evidence to support institutional policies and pedagogical guidance in the new phase of
digital transformation.
Addressing these gaps, the present study investigates the factors influencing
university students’ intention to use AI tools for learning in Vietnam. Based on TAM,
UTAUT, and TPB, the study examines five explanatory constructs - perceived ease of use
(PEOU), perceived usefulness (PU), social influence (SI), AI self-efficacy (AISE), and trust in
AI (TR) - with intention to use AI tools for learning (IU) as the dependent variable. By
doing so, the study seeks to provide empirical evidence that is relevant not only to
technology acceptance theory but also to the strategic management of AI-supported
learning in higher education.
2. Literature review, theoretical background, and research model
2.1. Literature review
Over the last few years, research on the acceptance and use of artificial intelligence
in higher education has expanded rapidly. Recent systematic reviews show that most
studies still rely on frameworks such as TAM and UTAUT, favor quantitative designs, and
emphasize the roles of usefulness, ease of use, and facilitating conditions in explaining
students’ acceptance of AI [3]. Similarly, Yang [4], in a systematic review of ChatGPT
acceptance in higher education, notes that the literature still needs more longitudinal
evidence, broader cultural diversity, and clearer accounts of how contextual conditions
shape actual usage.
Within the Vietnamese context, Maheshwari [5] finds that perceived ease of use
directly influences students’ intention to adopt ChatGPT, whereas perceived usefulness
operates indirectly through mechanisms such as personalization and interactivity. This
suggests that at the early stage of AI adoption, convenience and smooth user experience
may play an especially important role. Extending the focus from initial adoption to
continuance intention, Duong [6] reports that students’ continuance intention to use
ChatGPT for learning is shaped by both attitude and trust. In that study, trust emerges as
a key mechanism once AI use moves beyond the novelty stage.

702