AI-Driven Strategies in Strategic Communication: Understanding University Students’ Attitudes

Abstract

Introduction: Artificial intelligence (AI) is rapidly reshaping higher education, making it essential to understand students’ acceptance of AI-based learning tools. This study extends the Technology Acceptance Model by incorporating AI Familiarity, Self-Efficacy, Perceived Personalization, Institutional Support, and Ethical Concerns. Using survey data from 473 students at Hebei Academy of Fine Arts and Structural Equation Modeling, the research tests eight hypothesized relationships. The goal is to clarify how AI-related factors influence Perceived Usefulness, Perceived Ease of Use, and students’ attitudes toward AI adoption.

Objectives: This study examines university students’ acceptance of AI-based learning tools by extending the Technology Acceptance Model (TAM) with additional constructs relevant to AI use in education, including AI Familiarity, Self-Efficacy, Perceived Personalization, Institutional Support, and Perceived Ethical Concerns.

Methods: A cross-sectional quantitative survey was conducted with 473 students from Hebei Academy of Fine Arts in China. Structural Equation Modeling (SEM) was employed to assess the measurement and structural models, evaluate construct validity, and test eight hypothesized relationships within the extended TAM framework.

Results: Seven of the eight hypotheses were supported. AI Familiarity and Self-Efficacy significantly increased Perceived Ease of Use, while Perceived Personalization and Institutional Support positively influenced Perceived Usefulness. Perceived Usefulness and Perceived Ease of Use strongly predicted students’ attitudes toward AI adoption. Perceived Ethical Concerns did not significantly affect attitudes. The model demonstrated strong explanatory power, with ATT (R² = 0.768) being the most strongly predicted construct.

Conclusions: Findings show that AI Familiarity and Self-Efficacy improve Perceived Ease of Use, while Personalization and Institutional Support enhance Perceived Usefulness; Ethical Concerns show no significant effect. Perceived Usefulness and Perceived Ease of Use remain key predictors of attitudes, and the model demonstrates strong explanatory power (R² = 0.768). Although limited by a single-institution sample and self-reported data, the study highlights directions for broader future research. Overall, it enriches TAM literature in the AI education context and offers practical guidance for fostering supportive and responsible AI-enhanced learning environments.