We analyzed 74 OECD documents to identify core competencies of AI literacy, using the KSA framework—Knowledge, Skills, and Attitudes as the analytical basis. SBERT embedding and semantic-based clustering techniques were applied. The analysis identif...
We analyzed 74 OECD documents to identify core competencies of AI literacy, using the KSA framework—Knowledge, Skills, and Attitudes as the analytical basis. SBERT embedding and semantic-based clustering techniques were applied. The analysis identified 27 core competencies, with 9 competencies in each KSA category. In the knowledge domain, we identified competencies such as ‘Open Knowledge Sharing’ and ‘Comprehensive AI Educational Competency.’ In the skills domain, ‘Technology-based Problem Solving’ and ‘Creative Critical Thinking’ were confirmed. In the attitudes domain, competencies included ‘Empathetic Social Awareness’ and ‘Ethical Technology Consideration.’ We also proposed evaluation elements for each category; AI information interpretation and ethical judgment abilities for the knowledge domain; problem-solving and critical thinking abilities for the skills domain; and empathy and ethical responsibility for the attitudes domain. This study provides empirical criteria for AI education and policy development, supporting curriculum design and the construction of competency assessment systems.