Indonesia's Draft AI Ethics Regulation Adopts Three-Tier Risk-Based Approach
Kemkomdigi confirms the AI ethics presidential regulation classifies systems into low, high and unacceptable risk, with AI content labelling rules under study.
Indonesia's Ministry of Communication and Digital Affairs (Kemkomdigi) confirmed that the draft presidential regulation on AI ethics will adopt a risk-based approach with three categories: low, high, and unacceptable risk. Systems that process specific personal data fall into the high-risk tier and will face sector-specific requirements, with each ministry and agency setting its own risk standards. The draft also mandates incident reporting mechanisms and self-assessment of AI maturity.
Separately, the ministry is studying a ministerial regulation that would require AI-generated products and content to carry labels, using metadata-based authentication and provenance. The work sits within the National AI Roadmap 2026-2029, which involves more than 400 institutions, and officials have emphasised transparency and algorithm accountability as guiding principles. The presidential regulation is still being processed and no criminal sanctions or completion timeline have been announced.