Image source: https://www.worldbank.org/en/topic/edutech/publication/digital-pathways-education-enabling-learning-impact

Author: Hosea Immanuel Latumahina, Achmed Faiz Siregar Editor: Syaifa Tania, Iradat Wirid, Adya Nisita

Indonesia is no longer debating whether digital technology and artificial intelligence (AI) will enter education. They are already there. Around 89% of Indonesian children access the internet daily, spending an average of 5.4 hours online.[1] At the higher-education level, AI adoption is even more pronounced: 95% of Indonesian university students use AI to support their studies, while 86% use generative AI specifically for academic assignments.[2] Public familiarity is also substantial, with 83.6% of Indonesians reporting familiarity with AI technologies.[3] The direction of travel is therefore clear. Indonesia is no longer primarily facing an adoption problem, but is facing a governance problem.

High adoption, however, should not be mistaken for high readiness. Only 37.5% of children surveyed had received information about how to stay safe online, while 42% had experienced something online that made them uncomfortable or scared.[4] These vulnerabilities coexist with persistent weaknesses in learning outcomes. Indonesia’s performance in 2025 was 364 in mathematics, 365 in reading, and 389 in science.[5] However, these PISA scores remain low compared to the average scores of OECD countries. The percentage of students scoring above the baseline score is 37% in science, which is under the average for OECD countries (74%); 17% in mathematics, which is under the average for OECD countries (65%); and 27% in reading, which is under the average for OECD countries (69%).[6] The contradiction is difficult to ignore that children and students are becoming increasingly connected to digital technologies, while the human, institutional and regulatory capacities needed to make that connectivity safe, meaningful and equitable remain uneven.

The Paradox of Digital Technology and AI in Education

The paradox becomes even sharper with generative AI. It can accelerate writing, information retrieval, translation and content production, potentially reducing workloads for both students and educators. Yet greater efficiency does not automatically translate into deeper learning. Recent experimental evidence has raised concerns about changes in cognitive engagement and memory-related outcomes when participants relied on LLM assistance for writing tasks.[7] While such findings remain preliminary and should not be interpreted as definitive evidence of long-term cognitive harm, they point to an important question: what should educational technology ultimately optimize between faster task completion or better learning?

At the same time, AI adoption is not simply a matter of student behaviour. Teachers are also turning to AI to address practical demands within the classroom. A national survey found that 72.2% of elementary and secondary school teachers needed AI assistance to develop assessment questions, illustrating how rapidly AI is becoming embedded in everyday teaching practices.[8] This suggests that governance must account not only for how students use AI, but also for the institutional pressures that make AI attractive to educators.

More importantly, AI-enabled learning increasingly extends beyond the classroom. Students tend to use generative AI more frequently for school assignments at home than during classroom activities.[9] Their interaction with AI is consequently shaped not only by teachers and schools, but also by parents, caregivers, peers and digital platforms. The unit of governance, therefore, should not be the classroom alone, but the wider learning ecosystem surrounding the child.

From Tech to Non-Tech: Uneven Risks and Challenges Remain High

The implications of digitalisation in education extend far beyond technological access and instructional efficiency. They intersect with child protection, academic integrity, privacy and data governance, digital wellness, and institutional accountability. Child protection has become increasingly intertwined with digital learning. Greater connectivity can expose children to harmful content, cyberbullying, grooming, and online sexual exploitation and abuse. The emergence of AI-generated abusive material further complicates an already difficult protection landscape. These risks cannot be treated as peripheral to education because safety, participation and learning are deeply interconnected.[10]

Academic integrity presents another challenge. Generative AI increasingly blurs the boundary between legitimate assistance and substantive production. A student may use AI to brainstorm, translate, edit, restructure or generate an entire assignment, while existing academic rules may not clearly distinguish between these practices. Without shared expectations regarding disclosure, attribution and acceptable use, simply prohibiting AI may do little to develop responsible technological judgement.

Privacy and data governance constitute an equally important concern. Educational digitalisation generates increasingly extensive flows of personal data, including information relating to children. Questions of consent, data minimisation, retention, secondary use and institutional accountability therefore become central to responsible AI adoption. A human-centred approach requires not only safer AI outputs, but also meaningful safeguards over the data underpinning those systems.[11]

Digital wellness adds another dimension. Excessive technology use, reduced offline interaction and concerns over weakened reflective engagement suggest that educational success cannot be measured solely through technological access or productivity. Digital learning should leave sufficient space for physical activity, social interaction, creativity, character development and mental well-being.

These risks are already addressed, at least in part, through multiple legal instruments, including the Electronic Information and Transactions framework, Child Protection Law, National Education System Law, Personal Data Protection Law, PP No. 17/2025, and Presidential Regulation No. 87/2025.[12] The challenge, therefore, is not simply a lack of regulation. It is the lack of sufficient coherence across regulatory boundaries.

Multilayer Gap: Regulatory, Governance, and Implementation

Three interconnected gaps emerge from this fragmentation. First, there is a regulatory gap. Existing laws provide important foundations, but relevant provisions remain dispersed across sectors. Existing regulations and guidelines across line ministries provide a foundation for responsible AI use in education, but their implementation and integration remain uneven. Greater coherence is needed in translating these frameworks into operational standards for ethical AI in pedagogy, appropriate GenAI use, academic integrity, child-centred safeguards and digital wellness. The pace of technological development makes this challenge even more pronounced, as emerging AI capabilities can evolve faster than regulatory frameworks can adapt.[13]

Second, there is a governance gap. Education authorities, digital regulators, child-protection institutions and families each hold different parts of the responsibility. Yet cross-cutting problems rarely respect these institutional boundaries. Without clear mechanisms for coordination and role division, fragmented mandates can create a paradox in which everyone has part of the responsibility, but no actor fully owns the intersection.

Third, there is an implementation gap. Infrastructure, teacher AI literacy and digital parenting capacity remain uneven across geographical and socio-economic contexts. The same technology can therefore generate very different educational outcomes depending on where and how it is introduced. Without targeted capacity-building and differentiated implementation, digital transformation risks reproducing existing inequalities rather than reducing them.

Beyond Adoption: How the Government Responds?

The multidimensional character of these risks makes purely sectoral responses inherently partial. Child online safety is simultaneously an education, child-protection and digital-governance issue. Academic integrity involves pedagogy, institutional governance and technology. Digital wellness intersects with education, family practices and broader well-being. When the problem crosses institutional boundaries, governance must do the same.

This is where the Joint Ministerial Decree issued by seven ministers (Surat Keputusan Bersama Tujuh Menteri/SKB) 2026 becomes particularly significant. Signed on 12 March 2026, the SKB establishes guidance for the utilisation and learning of digital technology and AI across formal, non-formal and informal education pathways. Rather than merely introducing another layer of regulation, its potential value lies in connecting existing mandates and translating broad principles into coordinated operational action.

The rationale is particularly compelling given the asymmetry between technological and regulatory change. AI capabilities can evolve within months, whereas comprehensive legislation requires lengthy processes of formulation, consultation and enactment. A joint ministerial instrument can therefore function as a policy bridge to translating existing legal principles into shared operational expectations, clarifying institutional responsibilities and enabling coordinated responses without having to wait for an entirely new legislative architecture.[14]

Its importance lies in four functions: bridging fragmented policy frameworks; clarifying institutional roles; establishing shared principles for human-centred and ethical AI use; and legitimising collaborative action in areas such as teacher capacity-building, family digital literacy, child protection and monitoring.

Beyond the Classroom

The SKB’s broader significance is its recognition that education does not take place exclusively within classrooms. Children learn through interactions with teachers, parents, caregivers, peers, communities and digital platforms. Governance must therefore encompass the wider ecosystem surrounding the learner. Accordingly, the SKB places human-centred principles and child protection at the heart of digital technology and AI utilisation; promotes balanced digital engagement alongside physical and mental well-being; strengthens the roles of parents, guardians and educators through guidance, supervision and appropriate limitations; and emphasises digital and AI literacy grounded in ethics and academic integrity.[15]

In this context, the SKB 7 Menteri 2026 represents an important shift. from regulating technology within individual sectors towards coordinating responsibility across the ecosystem in which children learn and live. The objective of digital transformation should not be to make education more technological. It should be to ensure that technology remains human-centred, safe, ethical and genuinely conducive to learning. Going beyond adoption means moving from asking how much technology education can absorb to asking how well society can govern the world that technology is creating.

1. UNICEF Indonesia (2023), Online Knowledge and Practice of Children in Indonesia: A Baseline Study 2023, Jakarta: UNICEF Indonesia, available at: https://www.unicef.org/indonesia/child-protection/reports/online-knowledge-and-practice-children-indonesia-baseline-study-2023↑
2. Chegg (2025), Global Student Survey 2025, New York: Chegg, available at: https://www.chegg.org/global-student-survey-2025 ↑
3. Katadata Insight Center (2025), Kedaulatan AI untuk Memberdayakan Indonesia, Jakarta: Katadata Insight Center, available at: https://databoks.katadata.co.id/publikasi/2025/01/23/kedaulatan-ai-untuk-memberdayakan-indonesia ↑
4. UNICEF Indonesia (2023), Online Knowledge and Practice of Children in Indonesia: A Baseline Study 2023, Jakarta: UNICEF Indonesia, available at: https://www.unicef.org/indonesia/child-protection/reports/online-knowledge-and-practice-children-indonesia-baseline-study-2023 ↑
5. OECD (2026), PISA 2025 Results (Volume I) – Country Notes: Indonesia, Paris: OECD Publishing, available at: https://www.oecd.org/content/dam/oecd/en/publications/reports/2026/09/pisa-2025-results-volume-i-country-notes_88d1164e/indonesia_20c83120/9c880f8a-en.pdf ↑
6. Ibid. ↑
7. Turnitin (2025), ‘Turnitin releases new findings and key insights from focused survey on impact of AI in education’, Turnitin, available at: https://www.turnitin.com/press/turnitin-releases-new-findings-and-key-insights-from-focused-survey-on-impact-of-ai-in-education ↑
8. OECD (2023), PISA 2022 Results (Volume I and II): Country Notes – Indonesia, Paris: OECD Publishing, available at: https://www.oecd.org/en/publications/pisa-2022-results-volume-i-and-ii-country-notes_ed6fbcc5-en/indonesia_c2e1ae0e-en.html↑
9. Kosmyna, N., Hauptmann, E., Yuan, Y.T., Situ, J., Liao, X.H., Beresnitzky, A.A., Braunstein, I. and Maes, P. (2025), ‘Your Brain on ChatGPT: Accumulation of Cognitive Debt when Using an AI Assistant for Essay Writing Task’, arXiv, 2506.08872, doi:10.48550/arXiv.2506.08872 ↑
10. UNICEF Indonesia (2025), Online Knowledge and Practice of Children in Indonesia: A Baseline Study 2023, Jakarta: UNICEF Indonesia, available at: https://www.unicef.org/indonesia/child-protection/reports/online-knowledge-and-practice-children-indonesia-baseline-study-2023 ↑
11. Aisyah, N., Al Kautsar, M.D., Hidayat, A. and Koto, F. (2026), ‘Grounding AI-in-Education Development in Teachers’ Voices: Findings from a National Survey in Indonesia’, arXiv, 2604.01630, doi:10.48550/arXiv.2604.01630 ↑
12. Borgonovi, F., Bastagli, F., Ochojska, M. and Piumatti, G. (2025), AI Adoption in the Education System, OECD Artificial Intelligence Papers, Paris: OECD Publishing, doi:10.1787/69bd0a4a-en. ↑
13. Miao, F. and Holmes, W. (2023), Guidance for Generative AI in Education and Research, Paris: UNESCO, available at: https://www.unesco.org/en/articles/guidance-generative-ai-education-and-research ↑
14. Miao, F. and Holmes, W. (2023), Guidance for Generative AI in Education and Research, Paris: UNESCO, available at: https://www.unesco.org/en/articles/guidance-generative-ai-education-and-research ↑
15. Republik Indonesia (2026), Keputusan Bersama Menteri Dalam Negeri, Menteri Agama, Menteri Pendidikan Dasar dan Menengah, Menteri Pendidikan Tinggi, Sains, dan Teknologi, Menteri Komunikasi dan Digital, Menteri Kependudukan dan Pembangunan Keluarga/Kepala Badan Kependudukan dan Keluarga Berencana Nasional, dan Menteri Pemberdayaan Perempuan dan Perlindungan Anak tentang Pedoman Pemanfaatan dan Pembelajaran Teknologi Digital dan Kecerdasan Artifisial pada Jalur Pendidikan Formal, Nonformal, dan Informal ↑