As the largest economy in Southeast Asia, Indonesia faces significant challenges and opportunities in adopting and regulating artificial intelligence (AI). With its large population and continuously expanding economy, the potential for AI to drive change across various sectors is substantial. However, policymakers and stakeholders must navigate AI policies carefully to ensure that the benefits are felt across all layers of society, especially in Indonesia, where the diverse and culturally rich society has varied values and norms.

To address some of these challenges, the “FAIR Forward — Artificial Intelligence for All,” a global initiative in 7 partner countries across Africa and Asia on behalf of the German Federal Ministry for Economic Cooperation and Development (BMZ), GIZ together with the Ministry of National Development Planning/BAPPENAS of the Republic of Indonesia, has organized capacity-building training in the field of ethical AI Technology for Policymakers. This program, which runs from November 2023 to March 2024, has been attended by various stakeholders, ranging from Ministerial Institutions, Representatives of Regional Governments and Non-Governmental Organizations (NGOs) to academics. The training sessions have also featured speakers from ministries, NGOs, research bodies, and industry players.

Social Impact Assessment of AI

AI has the potential to contribute positively to social and economic development. However, its use can pose risks, such as job losses due to automation, bias and discrimination, and privacy and data security issues. In a presentation by Maria Hattya Karienova from EngageMedia in the Social Impact Assessments during the AI Policymakers Training, she delved deeper into a journal written by Natalie A. Smuha, “Beyond the individual: governing AI’s societal harm”, which discusses three types of harms that can arise from the use of AI: individual harm, collective harm, and social harm. Individual harms, such as the loss of privacy, are often discussed in the context of AI. However, Smuha invites us to look further. Collective harms, such as societal polarization, and social harms, such as inequality, are broader impacts that also need attention in policy-making.

\"\"/ — Sources: Documentation of the AI Policymakers Training, Bali, 4–5 March 2024

Social harms are often overlooked but cannot merely be reduced to the two previous types of harm. Smuha emphasizes the need for a shift in the perspective of AI regulation from a focus on the individual to a broader focus on the effects of AI on society as a whole. By understanding that AI impacts not only individuals but also our social structures, it is hoped that policymakers can formulate more comprehensive regulatory approaches to guide the ethical and responsible development and implementation of AI. Therefore, social impact assessments have become crucial in developing AI policies in Indonesia.

Risk-Based Approach in AI Policy

A risk-based approach enables policymakers to identify and manage potential risks associated with AI. This includes establishing a regulatory framework that is flexible yet robust enough to protect individual rights and foster innovation. The Indonesian National Strategy on Artificial Intelligence, or Stranas KA, is an example of a risk-based approach adopted by the Indonesian government. As presented by one of the speakers, Matheace Ramaputra, an AI and technology policy consultant, this strategy outlines the national policy direction designed until 2045 to guide the development and implementation of AI technologies in Indonesia. This strategy includes focus areas and priority fields of AI technology as references for ministries, agencies, local governments, and other stakeholders in carrying out activities in the field of AI.

The Indonesian National Strategy on Artificial Intelligence was formulated due to discussions by the National Strategy Formulation Working Group for Artificial Intelligence, established by the Agency for the Assessment and Application of Technology (BPPT). This group is comprised of experts representing the Quadruple Helix model, which involves the government, universities, industry, and community in AI. This strategic document also aims to serve as a reference in regulations and legislation related to AI technology in Indonesia. Moreover, this strategy has incorporated directions from stakeholders and is expected to receive input from the Indonesian public through public testing.

Regulatory Considerations

Existing regulations must adapt to the rapid advancements in AI technology. Indonesia has taken significant steps by introducing legally binding AI regulations to transition from soft ethical guidelines to lawfully enforceable rules.

\"\"/ — Ari Juliano Gema — Pengacara HKI
Sources: Documentation of the AI Policymakers Training, Bali, 4–5 March 2024

Intellectual Property Rights (IPR) considerations are also essential to AI policy. With the increasing development of generative AI such as Chat GPT, Dal-E, Copilot, and others that can create works in seconds, whether written, visual, or auditory, a unique problem arises in making a job. Regarding the creation of a work, Ari Juliano, an expert and lawyer in IP, stated during the closing event of FAIR Forward Indonesia, “The active involvement of a person in directing the work of AI in realizing and working on the design of their creation, and in the process of correcting or revising the output from the AI, must be considered.” He added, “Considering that the training data used in developing AI can come from various sources, the output of the AI can have distinctive and personal characteristics from various sources that are its training data. Suppose the output of the AI does not have distinctive and personal characteristics closely associated with the person designing the creation. In that case, that person does not meet the qualifications as a Creator.”

Facing these challenges, he offered several recommendations for policymakers, including: Requiring operators of generative AI as electronic system operators to register any generative AI applications offered in Indonesia with the Ministry of Communication and Information Technology (Kemenkominfo) and to disclose information about the training data sources used for developing their AI applications, including license or permission information if the training data involves data protected by intellectual property rights (IPR); and requiring developers/operators of generative AI to give credit or cite the sources of the training data that come from IPR-protected data in the outputs of the generative AI applications used by consumers.

Considerations for Disability

AI offers the opportunity to create a more inclusive environment that can accommodate the needs of individuals with disabilities. However, if poorly managed, AI can also exacerbate the challenges faced by people with disabilities in the labor market. I Made Prasetya Wiguna Mahayasa, Chief Executive Officer of PT Mahayasa Teknologi Nusantara or Annika Linden Center and People with Disabilities representative of DNetwork (Jaringan Kerja Disabilitas) echoed this during the closing of the FAIR Forward: Artificial Intelligence for All — AI Policymakers Training in Bali last March. As a technology practitioner and a person with multiple disabilities, Wiguna observed various solutions offered for people with disabilities. However, he noted that many of his peers are not well-socialized with these solutions and are not involved in finding these solutions. He emphasized the importance of socialization through expanding knowledge and providing mentors for people with disabilities utilizing AI technology. Therefore, policymakers must ensure that AI is designed with accessibility and inclusiveness.

\"\"/ — I Made Prasetya Wiguna Mahayasa, Chief Executive Officer PT. Mahayasa Teknologi Nusantara
Image Source: Dokumentasi *AI Policymakers Training*, Bali, 4–5 Maret 2024

Conclusion

Navigating AI policy in Indonesia requires a comprehensive approach that considers technical and economic aspects as well as social impacts, risks, regulations, intellectual property rights, and disabilities. With the right approach, AI can catalyze inclusive growth and development in Indonesia.

Reference: