Indonesian Ethics On Artificial Intelligence – Presidential Regulation Draft On Ethics Of Artificial Intelligence Dated 27 January 2026

The Government of Indonesia is preparing a draft Presidential Regulation on Ethics of Artificial Intelligence dated 27 January 2026 (“Draft AI Ethics Regulation”). If enacted into regulation in its current form, it would introduce a national ethics framework for the development, deployment, and use of Artificial Intelligence (“AI”) in Indonesia.

The Draft AI Ethics Regulation is particularly relevant for business actors in the AI ecosystem, which include sector actors, data providers, AI developers, AI system providers, and AI system operators (collectively, “AI Business Actors”). In addition to the Government and AI users, AI Business Actors are expected to play an important role in ensuring that AI systems are developed, operated, and used in accordance with ethical principles throughout the AI lifecycle.

This article is prepared based on the draft text of the Draft AI Ethics Regulation, including its Appendix. The key points are set out below:

1.          Risk Identification and Mitigation of the Use of AI

The Draft AI Ethics Regulation introduces a risk-based approach to the development, operation, and use of AI. Article 6 classifies AI-related risks into 3 categories, namely: (i) unacceptable risk; (ii) high risk; and (iii) low risk.

1. Unacceptable Risk: Under Article 7 (1) of the Draft AI Ethics Regulation, unacceptable risk refers to the use of AI that threatens the users’ safety and human rights. The Appendix of the Draft AI Ethics Regulation provides examples of AI uses that may fall within the unacceptable risk category, including:
   1. Real time face recognition in public spaces to monitor public movement without a legal basis, which may violate personal data protection and privacy, freedom of expression, and human rights protection.
   2. Public social scoring by combining specific personal data to determine access to certain services or facilities, which may result in discrimination, undermine individual autonomy, and violate human rights.
2. High Risk: Under Article 7 (2) of the Draft AI Ethics Regulation, high-risk AI includes AI that uses specific personal data; and AI that may have a significant impact on human rights, safety, essential public services, employment, health, education, law enforcement, financial access, or other sensitive areas.

The Draft AI Ethics Regulation refers to specific personal data which includes: (i) health data and information; (ii) biometric data; (iii) genetic data; (iv) criminal records; (v) children’s data; (vi) personal financial data; and (vii) other data as determined by the laws and regulations. These categories are broadly consistent with the categories of specific personal data under Law No. 27 of 2022 on Personal Data Protection.

For AI Business Actors, this means that AI systems involving specific personal data or decisions affecting sensitive rights and interests should be subject to more robust governance. This may include data protection review, bias testing, human oversight, audit trail, explainability measures, security controls, and incident response procedures.

3. Low Risk: Under Article 7 (3) of the Draft AI Ethics Regulation, low-risk AI refers to AI that creates minimal or no threat to human rights or public safety. The Appendix of the Draft AI Ethics Regulation gives examples of low-risk AI, including:
   1. An AI-powered public information service chatbot that provides general information, such as office hours, administrative requirements, or licensing applications procedures.
   2. AI-based product recommendation systems used by e-commerce platforms, provided that they do not have a significant impact on user safety, human rights, or access to essential services.

Although low-risk AI is subject to a lighter risk profile, AI Business Actors should still ensure that users are not misled, personal data is processed lawfully, and the AI system does not produce harmful, discriminatory, or inaccurate outputs.

2.          Monitoring and Evaluation of AI Use

The Draft AI Ethics Regulation also introduces a monitoring and evaluation mechanism for AI ethics. Article 9 (1) of the Draft AI Ethics Regulation, ministries, institutions, and sector actors are required to conduct monitoring and evaluation of the implementation of AI ethics.

Further, Article 9 (2) of the Draft AI Ethics Regulation provides that such monitoring and evaluation may be conducted through the following mechanisms:

1. Self-assessment: AI Business Actors are expected to conduct self-assessment of their AI systems. Based on the Appendix of the Draft AI Ethics Regulation, this may include an assessment of bias detection and mitigation, decision-making accountability, privacy protection, social risks, and other governance measures.
2. Incidental reporting: The Draft AI Ethics Regulation contemplates incident reporting, where complaints or alleged ethics violations may be received and handled by the relevant ministry or institution, and may then be forwarded to the relevant AI Business Actor for follow up.
3. Sectoral reporting: Sectoral reporting is to be conducted by ministries or institutions on a regular basis, allowing sectoral regulators to monitor AI implementation in areas such as health, finance, education, employment, public services, and other regulated sectors.
4. Integrated monitoring system: The Draft AI Ethics Regulation also provides an integrated monitoring system to be developed and operated by the relevant ministry to support the monitoring and evaluation of AI ethics implementation. At this stage, the draft does not provide detailed technical requirements for this system. Accordingly, further implementing rules or guidelines may be needed to clarify its scope, reporting format, responsible authority, data submission requirements, confidentiality treatment, and follow-up procedures.

As a note, under Article 10 of the Draft AI Ethics Regulation, the monitoring and evaluation provisions do not apply to AI use for: (i) strategic and sensitive matters of national defence and security; and (ii) non-commercial testing, development, and research, in each case, to the extent they are conducted in accordance with the prevailing laws and regulations.

3.          Implementation of AI Ethics

Article 11 (1) of the Draft AI Ethics Regulation mandates ministries, institutions, and sector actors to prepare the implementation of AI Ethics no later than 2 years after the regulation comes into effect. For AI Business Actors, this two-year period should be viewed as a preparation period to translate the ethical principles under the Draft AI Ethics Regulation into practical internal governance, including risk mapping, data governance review, accountability structures, safeguards for higher-risk AI systems, human oversight, documentation, and incident response procedures.

The Draft AI Ethics Regulation also recognises that AI ethics implementation will require a collaborative approach. Article 12 provides that AI ethics may be implemented through collaboration with relevant stakeholders, including the Government, industry players, academics, civil society organisations, and the media. This is important because AI governance is not only legal compliance, but also technical design, public trust, and social impact. In addition, Article 13 provides that the ministry responsible for communication and informatics affairs may facilitate coordination and implementation of AI ethics through the establishment of a National AI Ethics Forum. If implemented effectively, this forum may become a platform to align regulatory expectations, sectoral practices, and industry feedback as Indonesia’s AI governance framework continues to develop.

Key Takeaways

 The Draft AI Ethics Regulation is an early signal of how Indonesia may approach AI governance going forward. Rather than imposing a full licensing or sanctions regime at this stage, the draft sets out the basic expectations for responsible AI use, particularly through risk classification, safeguards, self-assessment, incident handling, and sectoral monitoring. For the public, this is relevant because AI systems can affect privacy, safety, equality, access to services, and other fundamental interests. For businesses, the message is that AI should not be treated only as a technology or efficiency tool. Companies that develop, provide, or use AI need to start building practical governance measures, including risk assessment, personal data compliance, human oversight, accountability processes, and incident response, especially where their AI systems involve specific personal data or decisions that may significantly affect individuals.

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.