A Comparative Analysis of the Factors Influencing EFL Students' and Instructors’ Acceptance and Adoption of AI-Assisted Language Learning: A UTAUT-Based Study on ChatGPT

Background: The emergence of artificial intelligence (AI) tools such as ChatGPT has transformed English as a Foreign Language (EFL) education by offering new approaches to language learning and teaching. However, the evidence base remains limited in three important respects: prior studies often homogenize students and instructors, report inconsistent findings on key acceptance determinants, and rarely examine AI adoption through the Unified Theory of Acceptance and Use of Technology 2 (UTAUT2) within culturally and pedagogically specific EFL contexts.

Purpose: To address inconsistent and fragmented evidence on AI adoption in language education, this study examines the determinants of ChatGPT adoption among Saudi EFL students and instructors using UTAUT2, and evaluates how demographic factors condition these relationships in a generative AI context.

Method: The study employed a quantitative cross-sectional survey with 345 participants (students and instructors) from four Saudi universities. Separate UTAUT2-based questionnaires were validated through pilot testing and confirmatory factor analysis. Multiple regression analyses were conducted separately for each group to examine determinant effects, with interaction terms used to test the moderating roles of gender, age, and experience.

Results: Across both groups, performance expectancy was the strongest common predictor of behavioral intention (students: β = 0.42; instructors: β = 0.40). Beyond this shared effect, the patterns diverged. For students, social influence (β = 0.30) and hedonic motivation (β = 0.22) also played a significant role, whereas effort expectancy and habit did not reach significance. For instructors, by contrast, effort expectancy (β = 0.11) and habit (β = 0.10) were significant alongside performance expectancy, while social influence and hedonic motivation were not. Facilitating conditions and price value were non-significant in both groups. The moderation analysis further indicated that gender, age, and experience affected selected relationships between predictors and behavioral intention, with these effects varying across students and instructors rather than following a uniform pattern.

Conclusion: Adoption of ChatGPT in Saudi EFL education depends on both shared and group-specific factors, with performance expectancy as the strongest common predictor. The results emphasize the need for culturally informed and targeted strategies that address demographic and experiential differences and provide theoretical and practical guidance for AI integration in language education.