Abstract

This cross-sectional study examines self-assessed AI literacy and perceived risk of AI replacement among 647 final-year students from private universities in Vietnam’s Central Highlands and Ho Chi Minh City. AI literacy was measured using the 12-item AILS, while replacement risk was assessed using a 10-item JIS measure. Data were analyzed using Welch tests, effect sizes, sensitivity analysis, equivalence testing, and general linear models. Mean scores were 4.78 (SD = 1.17) for AILS and 3.42 (SD = 0.85) for JIS. No significant differences were found by field of study, AI usage frequency, gender, or location. Equivalence tests supported similarity across gender and location, while adjusted models explained only 1.0% of AILS variance and 0.4% of JIS variance. The findings indicate that broad group characteristics have limited explanatory value. Student support should therefore prioritize meaningful AI experience, critical evaluation of AI outputs, and evidence-based career guidance.

Keywords: AI literacy, final-year students, generative artificial intelligence, perceived AI replacement risk, university students.