Thailand's Office of Insurance Commission published its AI Governance Guideline B.E. 2568 for insurance companies, setting supervisory expectations for the responsible, transparent and secure use of artificial intelligence while balancing innovation with consumer protection. The guideline follows a risk-based proportionality approach and emphasises that high-risk AI uses with direct impacts on policyholders should include human involvement in decision-making. The framework sets expectations across four areas: AI governance with clear structures, roles and accountability and staff capability on ethics and AI risk management; robustness and AI security across the system life cycle, including secure-by-design development, vulnerability management, logging, backup and recovery; transparency and explainability, including the ability to explain AI-driven decisions and outcomes where consumer rights and benefits are affected; and fairness and consumer protection, including controls over data quality to reduce data bias and pathways for consumers to request explanations or a human review of AI decisions. The Office noted that the guideline had already been subject to a public consultation process prior to publication.