Modeling behavioral intention toward generative AI use in higher education English language teaching

Introduction The rapid emergence of generative artificial intelligence (AI) presents new opportunities and challenges for English language teaching (ELT) in higher education, particularly in developing and resource-diverse contexts. Responding to limited empirical evidence from Philippine higher education, this study examines the behavioral intention of higher education English language teachers in Cebu to adopt generative AI in instructional practice. Guided by an integrated framework combining the Unified Theory of Acceptance and Use of Technology (UTAUT) and Expectancy–Value Theory (EVT), and extended with perceived knowledge of AI and perceived privacy concerns, the study adopts a predictive, theory-guided design. Methods Survey data were collected from 488 higher education teachers across urban and rural state universities and local colleges in Cebu and analyzed using Partial Least Squares Structural Equation Modeling (PLS-SEM). Results Results show that performance expectancy predicts perceived attainment value, effort expectancy predicts perceived utility value, and social influence shapes perceived attainment value, perceived intrinsic value, and perceived cost. Facilitating conditions predict both perceived utility value and perceived cost. Among downstream mechanisms, perceived attainment value and perceived utility value significantly predict perceived knowledge of AI, which emerges as the strongest direct predictor of behavioral intention. Perceived intrinsic value does not significantly predict perceived knowledge of AI, and perceived privacy concerns do not significantly predict behavioral intention after accounting for value-based and knowledge-based predictors. Moderation analysis indicates that age does not significantly moderate any structural relationships, while sex moderates only the relationship between effort expectancy and perceived utility value. Discussion Overall, the findings indicate that behavioral intention to adopt generative AI in higher education ELT is primarily driven by performance-related beliefs, value-based appraisals, and self-assessed AI competence, rather than by demographic characteristics or privacy concerns alone. The study contributes a context-sensitive, theory-integrated explanation of AI adoption intentions and offers implications for AI literacy, professional development, and responsible institutional integration in higher education English language teaching.