From policy frameworks to AI mentoring practice: A structured approach to responsible innovation in architectural education

Addressing the policy-to-practice gap in AI-supported architectural education, this study examines a structured “AI Mentoring Method” in alignment with King Abdulaziz University’s AI policy framework. Implemented in a senior undergraduate architectural research course across two consecutive semesters under different instructors, the method is organized into three stages: design, application, and evaluation, and systematically integrated across five course chapters. Generative AI tools were embedded through instructor-mediated tasks grounded in guided inquiry, scaffolding, and reflective practice. The study adopts an explanatory case study approach combining student satisfaction surveys, longitudinal quantitative assessment of Intended Learning Outcomes, and qualitative evidence from student work samples. Survey data were analyzed descriptively, while learning outcomes were compared across three semesters (pre-implementation and post-implementation under two instructors). Results indicate improved AI literacy, sustained learning gains, and strengthened value-based outcomes related to ethical awareness and academic responsibility. Variations between cohorts highlight the critical role of the instructor in shaping AI-supported learning, despite applying the same methodological framework. The study contributes a pedagogically grounded and policy-aligned model for responsible AI integration. While limited by a single-course context, the findings suggest that the AI Mentoring Method offers a transferable framework for structured, instructor-led AI adoption in design and research-based curricula.