Benchmarking Global Institutional AI Policies for Teaching and Learning: A Comparative Analysis for Higher Education Governance Enhancement

Abstract

This study benchmarks global institutional artificial intelligence (AI) policies for teaching and learning in higher education and uses these comparisons to improve an existing AI policy at a university in Northern Mindanao, Philippines. Although AI policies are spreading rapidly in higher education, few studies have systematically compared Philippine policies with global standards, particularly in teaching and learning. The study used qualitative benchmarking and document analysis. It reviewed AI policies and official guidance from ten top-ranked universities, chosen for their rankings, regional diversity, and public access. The analysis followed the OECD AI Principles and the NIST AI Risk Management Framework, focusing on ethics, governance, and risk management. The findings show that institutions share similar ethical commitments, especially to human-centered AI, academic integrity, transparency, and responsible use. However, global institutions have more advanced governance by putting ethical principles into practice with formal risk assessments, oversight, and regular policy reviews. In contrast, the local policy is mostly based on principles. The study concludes that while ethical standards for AI policy are becoming more global, the level of governance depends on each institution’s resources and regulations. It suggests a step-by-step policy improvement plan, starting with clear ethics, then adding risk mapping, governance structures, and ongoing review. The benchmarking framework offers a practical guide for universities aiming to improve AI governance in teaching and learning.