In December 2023, Indonesia saw the issuance of two sets of ethical guidelines on the use of artificial intelligence. The first was Ministry of Communication and Informatics (MOCI) Circular Letter No. 9 of 2023 on AI Ethical Guidelines, issued on 19 December 2023 (the MOCI AI Circular Letter). The second was Financial Services Authority (OJK) Ethical Guidelines on Responsible and Trustworthy AI in the Financial Technology (fintech) Industry, published on the OJK website on 4 December 2023 (the OJK AI Guidelines).
In this article, we will review the positions taken by MOCI and OJK on AI in Indonesia, and compare them with those taken by the European Union, Australia and Singapore.
MOCI AI Circular Letter
Significantly, the MOCI AI Circular Letter is addressed not only to business players with Indonesian Business Classification (KBLI) Code 62015 (AI-based programming activities), but also to all public and private electronic system operators engaging in AI-based programming activities. This is in contrast to the draft version, which was only for business players with KBLI Code 62015.
Under MOCI Regulation No. 3 of 2021 on Business Activity and Product Standards in Implementing Risk-Based Business Licensing in the Postal, Telecommunications and Electronic Systems and Transactions Sectors, business players with KBLI Code 62015 must put in place and implement internal company policies on data and AI ethics. The MOCI AI Circular Letter provides guidance for these businesses in preparing their company policies.
The MOCI AI Circular Letter highlights the benefits of AI for increasing productivity, optimising business processes, and providing more personalised services to customers in various sectors. These include the creative industry sector, which is embracing AI for content creation on social media, the health sector, which is incorporating AI to achieve more accurate medical diagnoses, and the education sector, which is integrating AI to support learning and research activities.
With the MOCI AI Circular Letter, Indonesia has kickstarted its AI regulatory journey to implement the 2020–2045 National Strategy for Artificial Intelligence, though this is still at an early stage compared to other countries.
At the time of writing, the European Union (EU) was in the process of finalising its draft “EU AI Act” which proposes to regulate AI applications across the region based on different risk tiers. Meanwhile, Australia released its AI ethics framework back in 2019 and is currently exploring options to regulate high-risk AI applications, while Singapore released model AI governance frameworks and toolkits in 2020.
Definition of AI
The MOCI AI Circular Letter defines AI as a form of programming on a computer device in carrying out precise data processing (pemrosesan) and/or data analysis (pengolahan), while the implementation of AI is described as activities related to AI research, product development, marketing and use, including consulting, analysis and programming. This is a broader definition of AI than that used by the EU, Australia and Singapore (see Definitions of AI around the world).
The MOCI AI Circular Letter stipulates that implementation of AI should observe various ethical values, which are broadly similar to those found in other jurisdictions, albeit with slightly different terminology. However, while Indonesia, the EU and Australia offer valuable perspectives on how to build a responsible and ethical future for AI, the ethics principles in Singapore appear more focused and practical.
The accompanying table compares the ethical values applied in each jurisdiction.
The MOCI AI Circular Letter sets out certain responsibilities, including ensuring that AI is not implemented as the sole determinant of policies and/or decisions concerning humanity, preventing the occurrence of racism and avoiding any action that harms people, among various other obligations.
OJK AI Guidelines
The OJK AI Guidelines apply to all fintech players in Indonesia. They acknowledge that AI and machine learning can improve the efficiency of business processes and the speed of financial services transactions. However, the use of AI also gives rise to unprecedented risks. As such, the guidelines are intended as a code of conduct to guide fintech players and related parties to make sure their AI-based applications comply with the following principles:
- based on Pancasila5, eg, the use of AI should align with national interests and uphold ethical responsibilities based on Pancasila values;
- beneficial, eg, the use of AI-based applications should give added value to business operations and improves consumer welfare;
- fair and accountable, eg, do not give rise to discrimination based on consumers’ lack of awareness regarding the use of black box AI technology or datasets that are inappropriate;
- transparent and explicable, eg, applying a “human-on-the-loop” approach to ensure that fintech players have the knowledge and ability to control AI-based application processes, and can explain them to consumers; and
- robustness and security, eg, ensuring that a recovery mechanism is in place in case of a cyber-attack.
The above principles were identified by OJK following a study of several guidelines issued globally, including the Organisation for Economic Co-operation and Development (OECD) AI Principle and the National Institute of Standards and Technology AI Risk Management Framework.
In contrast, the EU does not have specific regulations on the use of AI in fintech, though the preamble of the draft EU AI Act does mention credit scoring systems as an example of an application that could attract strict rules under the ‘high risk’ category.6
Likewise, while Australia does not have AI-specific regulations for fintech, there have been various industry-led guidelines such as the Australian Human Rights Commission’s “Guidance Resource on AI and Discrimination in Insurance”7, and existing technology-neutral laws that regulate aspects of the use of AI.
In Singapore, the Veritas Initiative aims to enable financial institutions to evaluate AI driven solutions against the principles of fairness, ethics, accountability and transparency (FEAT). Veritas is part of Singapore’s National AI Strategy with the Monetary Authority of Singapore, working alongside players in the financial industry, to strengthen internal governance around the application of AI and the management and use of data.
The issuance of these ethical guidelines on the use of AI in Indonesia is a significant proactive step towards the establishment of AI regulations here, laying the groundwork for the development of a comprehensive regulatory framework for AI in Indonesia.
At the time of writing, we understand that MOCI is proposing a comprehensive regulation on AI, while the National Research and Innovation Agency (BRIN) is drafting a presidential regulation on AI.
We also note that the Association of Southeast Asian Nations (ASEAN) has just issued a Guide on AI Governance and Ethics. We expect more developments in this space shortly, not just in Indonesia but also in the EU, Australia, Singapore and beyond.
2 EU’s “Ethic Guidelines for Trustworthy AI” (2019), which form the basis of the draft EU AI Act
3 Australia’s “Commonwealth AI Ethic Principles” (2019)
4 AI Verify, Singapore’s AI Governance Testing Framework and Toolkit
5 Pancasila is the official philosophical foundation of Indonesia, enshrined in Indonesia’s Constitution. It consists of five principles: (i) belief in one and 3only God; (ii) just and civilised humanity; (iii) the unity of Indonesia; (iv) a democratic life guided by wisdom in deliberation/representation; and (v) realising social justice for all the people of Indonesia.
6 Draft EU AI Act, Preamble paragraph 37