Governing Artificial Intelligence in Thailand's Higher Vocational Education

Abstract

Artificial intelligence (AI) is entering vocational education not only as a teaching aid but also as infrastructure for curriculum planning, student support, work-based learning coordination, resource allocation, and quality assurance. In Thailand, these applications must serve the competency orientation of the Higher Vocational Certificate (HVC, or Por Wor Sor), comply with education and data-protection requirements, and remain auditable within internal and external quality-assurance systems. This article develops a context-specific governance and management framework through an integrative policy review. The analysis synthesizes Thai legislation, the 2024 vocational qualification standards and HVC curriculum, the National Qualifications Framework, the 2024–2028 external quality-assurance framework, Thailand’s national AI strategy and governance guidance, and international evidence and standards available through 30 July 2026. The review finds that AI’s strongest near-term contribution is decision augmentation: improving the timeliness and coherence of evidence while preserving professional judgement. Benefits are plausible across labour-market intelligence, learner-risk detection, dual-training logistics, asset planning, and evidence management; however, educational and legal harms become more serious when systems influence admission, assessment, certification, discipline, financial support, or workplace placement. The proposed HVC-AI Governance and Management Framework integrates six components—public value, human authority, data stewardship, proportional risk control, lifecycle accountability, and continuous evaluation—with a three-tier use-case classification and a phased implementation roadmap. The article contributes a direct mapping between AI use cases and Thailand’s vocational quality-assurance expectations. Because national causal evidence remains limited, the framework should be implemented as a testable policy proposition, using controlled pilots, disaggregated outcome measures, stakeholder participation, and independent review rather than technology-led scaling.