This study investigates determinants of students’ actual use of artificial intelligence (AI) in learning by integrating Technology Acceptance Model (TAM), Theory of Planned Behavior (TPB), and Unified Theory of Acceptance and Use of Technology (UTAUT) and incorporating Technological Barriers (TB) from the Technology–Organization–Environment (TOE) framework. Using a quantitative survey of 455 undergraduates at major universities in Hanoi, Vietnam, we assessed reliability, conducted Exploratory Factor Analysis (EFA), correlations, and multiple linear regression. The model explains 59.2% of the variance in the Decision to Adopt AI for Learning Purposes (DAILP) (R = 0.769; R² = 0.592; DW = 1.906). All six hypothesized effects are significant (p 0.01): Social Influence (SI) (β = 0.276), Perceived Usefulness (PU) (β = 0.262), Perceived Behavioral Control (PBC) (β = 0.220), Performance Expectancy (PE) (β = 0.178), Perceived Ease of Use (PEOU) (β = 0.167), and TB (β = −0.178). Results highlight the primacy of social endorsement and perceived utility in driving AI adoption, while infrastructural and skill constraints suppress use. We discuss actionable implications for universities—strengthening peer/faculty norms, embedding course-aligned AI use-cases, building AI literacy, integrating tools with LMS, and reducing access frictions—alongside integrity safeguards. The study advances acceptance research by explaining the actual use of and contextualizing inhibitors that are salient to developing countries’ higher education.
Le Thi Kim Hoa (Fri,) studied this question.