Ethical concerns about generative AI in higher-education learning experiences: self-efficacy from students’ perspectives

GAI is becoming part of university learning, yet students must ethically make decisions. This mixed-methods study examines ethical concerns of GAI among tourism and hospitality students in Macao. Four focus groups with 28 undergraduate and postgraduate students identified four salient concerns: academic integrity, personal dependency, hallucination, and employability. These student-derived themes informed construct selection, measurement adaptation, and hypothesis development for a survey of 271 students. Owing to social cognition theory, the results showed that each concern was negatively associated with self-efficacy, which was positively associated with intention to use GAI. Self-efficacy also mediated each ethical concern-intention relationship. Personal dependency had the largest negative association with self-efficacy and the highest importance in the importance-performance map. This study shows how ethical appraisals can shape perceived capability and, in turn, willingness to use GAI. The findings support curriculum-based AI literacy, faculty development, transparent institutional governance, and industry collaboration that preserves human-centered professional competencies.