Indonesia Targets September 2025 Harmonization of AI Presidential Regulation, Backed by National Roadmap and Ethics Guidelines

Indonesia's AI Perpres set for harmonization by end of September 2025

The government plans to bring the draft presidential regulation (Perpres) on artificial intelligence to the harmonization phase by the end of September 2025. The Ministry of Communication and Digital Affairs will begin deliberations on legal principles to ensure the regulation fits with existing laws and avoids overlaps.

Harmonization is a formal process to align a draft with legal instruments across levels of regulation so it does not conflict with current rules. In parallel, the government is preparing a White Paper on the National AI Road Map and guidelines on security and safety for advanced digital technologies.

White Paper and safety guidelines

The White Paper has involved 443 contributors from government, academia, industry, communities, and media. Its goal is to support inclusive and responsible AI use in Indonesia and serve as a reference for strategies and policies.

The forthcoming safety guidelines will complement the White Paper and provide practical guardrails for development and deployment. Both documents are expected to feed into the AI Perpres.

Strategic direction

The government is building a national AI road map to address social and economic impacts while boosting global competitiveness. The stated vision is an ethical and accountable AI ecosystem that supports "Golden Indonesia 2045."

Leadership has also stressed the need to unify stakeholder direction on AI development at home. Ethics guidelines are being prepared to complement provisions introduced via a ministerial circular in 2023.

What government agencies can do now

• Appoint an AI focal point and set up a cross-functional working group (policy, legal, IT, procurement, risk, communications).
• Inventory current and planned AI use cases and vendors; flag high-impact areas (public benefits, social assistance, health, law enforcement).
• Create a basic AI risk and impact assessment template (purpose, datasets, model type, risks, mitigations, human oversight, exit options).
• Strengthen data governance: classification, consent basis, retention, access controls, logging, and data-sharing agreements.
• Update procurement clauses: data-use limits, IP and privacy terms, model documentation, evaluation rights, security obligations, incident reporting, service-levels.
• Define human-in-the-loop and appeal processes for automated decisions that affect rights or entitlements.
• Set evaluation protocols (accuracy, fairness, safety, security) and performance thresholds; schedule periodic re-testing.
• Establish security baselines for AI systems and supply chains, including adversarial testing and monitoring.
• Open channels for public information and complaints about agency AI use; publish clear notices for pilots and deployments.
• Plan capacity building for policy, legal, audit, and technical teams; prioritize practical training and certifications.

What to watch next

• Harmonization outcomes for the AI Perpres and any public consultation windows.
• Publication of the National AI Road Map White Paper.
• Release of security and safety guidelines and how they apply to public-sector use.
• Inter-agency coordination guidance for consistent adoption across central and regional bodies.

For reference on international standards, see the OECD AI Principles and the UNESCO Recommendation on the Ethics of AI.

Need practical upskilling for public-sector teams? Explore curated AI courses by job function at Complete AI Training or review relevant AI certifications.