Gertrude Grace N. Regaspe, Allen C. Laxa, Franco Eduardo E. Garcia, Louise Miguel F. Israel, Jinky P. Batiduan, John Christopher B. Mesana, Allan B. de Guzman
Artificial intelligence (AI) is increasingly embedded in higher education and tourism learning. This study used conjoint analysis to examine AI-use preferences among 416 tourism students in a metropolitan Philippine higher education context. Sixteen estimation profiles and two holdout profiles were generated through an orthogonal design and ranked through a three-stage sorting procedure. The estimation profiles showed moderate agreement, with Pearson’s r = .717 (p < .001) and Kendall’s τ = .483 (p < .05); however, the holdout profiles showed negative predictive validity (τ = −1.000). Purpose of use (30.97%), payment method (22.80%), and prompting style (22.61%) were the most influential attributes. Students favored examinations, projects, and idea generation; monthly payment; subscription-based access; manual setup; and detailed prompts. The findings provide context-bound evidence for assessment design, AI literacy, access planning, and curriculum development, but the failed holdout validation requires cautious interpretation and replication.