Author: Balma Bahira AdzkiaEditor: Achmed Faiz Yudha Siregar
Since its emergence in 2022, ChatGPT became the gateway to global AI discourse, including in Indonesia. It was followed by a range of other AI-based applications with diverse functions—productivity, creativity, and even automation—requiring today’s society to become increasingly relevant and to adopt them into everyday life.
However, the way this technology has been popularized often emphasizes its sophistication more than its consequences. In Indonesia, this situation is becoming increasingly complex[1]. On one hand, AI adoption keeps being pushed forward. On the other hand, its basic foundations are not yet fully in place: a declining digital literacy index, uneven digital inclusivity[2], and AI regulation that is only set to be passed this year.
As a student learning to build AI systems, while also witnessing how this technology is communicated outside the academic sphere, I see a fairly concerning gap between the two. AI is positioned as an instant answer carrying a go-to label, “cutting-edge”, rather than as a complex system with limitations and risks.
The Empty Loop: When Adoption Outpaces Readiness
In the context of AI model development, human-in-the-loop (HITL) means involving humans (usually experts) as supervisors, approvers, or correctors to ensure that AI’s final decisions remain accurate, ethical, and safe[3]. If this principle is elaborated into the everyday-use ecosystem, the “human” in question is no longer just the expert behind the screen but anyone who interacts with AI output and decides what to do with it. The question then shifts: “What kind of human should be inside that loop?”
Individual readiness for the AI era is not simply about the ability to operate applications but how a person verifies output, formulates the right questions, and recognises when AI is not a shortcut solution to a problem.
In Indonesia, this readiness is measured across four pillars: infrastructure & ecosystem, digital literacy, empowerment, and employment. The 2025 Indonesian Digital Society Index (IMDI) score released by Komdigi stands at 44.53, up one point from the previous year’s 43.34.
Yet the statistics by pillar paint a rather bleak picture. Among all the pillars that saw significant improvement, Digital Literacy experienced a drastic decline of nearly 10 points from 58.25 (2024) to 49.28 in 2025[4]. This is despite the fact that this pillar covers the criteria for successful information dissemination: skills in verifying information and assessing its credibility, awareness of digital security, and ethics in digital spaces.
In practice, these figures and charts seem not to factor into policy decisions at all. Even as the Digital Literacy score falls, the push to launch AI-based applications and services in the public sector keeps growing[5][6].
Digitalization and the use of AI appear politically rewarding because they offer solutions that appear fast and progressive. However, amid the existing reality, this action can lead to techno-solutionism, a state where technology is always believed to be the answer to a range of social problems[7].
An Unbalanced Narrative and Who Fills It
Techno-solutionism is not born from a single party’s decision. It is the product of misaligned incentives among the three actors most decisive in shaping the direction of Indonesia’s AI ecosystem: the AI community, government, and industry[8]. This misalignment drives an imbalance in the AI narrative, which currently highlights sophistication far more than risk.
The AI community—students, researchers, and practitioners—sits closest to how AI systems actually work. That position should make them a bridge between technological complexity and public understanding. Yet in digital spaces, the content this community produces most is centered on prompting tutorials and reviews of the latest tools; rarely does it touch on the discourse about responsibility, risk, or the limits of AI capability. Much of this content is even made by people who may themselves be little more than trend-watchers, rather than by those with a genuinely relevant background.
As a result, cases where AI-generated content causes social harm and personal damage rarely receive serious attention. For example, criminal fraud and defamation that manipulate the faces of public figures, celebrities[9], and even volunteers[10] using deepfake. Or in some cases, AI-reconstructed music uploaded freely to YouTube, racking up view counts that nearly rival the original work. In the absence of AI regulation, cases like these sit in a legal gray zone—lumped together with violations of the Electronic Information and Transactions (ITE) Law and the Copyright Law[11], and not infrequently going nowhere beyond the police report.
Beyond protecting citizens’ data and ensuring system accountability, a regulatory framework also determines how far Indonesia’s AI ecosystem can advance. AI is not merely a hype-driven trend to adopt, but a catalyst for building a more inclusive and equitable digital infrastructure for the common good[12]. When regulation lags behind adoption, the vision of a safe digital ecosystem drifts further out of reach.
Industry, meanwhile, has the clearest incentive of all: the market. A PwC survey[13] notes that AI adoption among Gen Z keeps rising. Yet the skills gap between generations remains significant. Growth in adoption that isn’t matched by investment in user literacy creates a large but fragile market; full of users who are proficient in operating tools without truly understanding the working mechanism and the consequences they cause.
Conclusion
If likened to a group assignment, it’s worth recognizing that the government, the AI community, and the industry each have their respective roles. I imagine all the actors in this ecosystem first have the same foundation regarding Indonesia’s readiness in adopting and facing the age of artificial intelligence. I am not pessimistic about the position of AI or how it will eventually be applied, so long as the people behind it have been prepared well ahead of time.
Starting from the fact that the four pillars underpinning a digitally literate society are still fragile, to the certainty and legal position of AI in the eyes of the law, to the technology consumerism the government often reaches for as a quick fix—these are the “homework” the government must resolve first. Beyond that, the government also needs to decide Indonesia’s position in contributing to the global AI world: will it remain a consumer forever, or is there an ambition that is not only political, but also democratic?
Becoming a “bridge” between experts and laypeople is one path the AI community can take toward this shared task. Technically, the people within this community have greater access and capacity to understand what is actually happening behind these intelligent machines. With a more balanced narrative between AI’s sophistication and its side effects, there is a greater chance AI will be seen as a benefit rather than a threat.
AI’s cutting-edge quality, often used as the flagship feature of the products built around it, is indeed a promising selling point. Even so, industry players also need to consider the actual urgency of such features for their target market. Cross-generational inclusivity is a crucial consideration here, as older generations tend to be more skeptical compared with newer generations who grew up alongside fast-evolving technology.
The role of humans within the AI ecosystem does not end with the ability to operate tools. Still, it extends to understanding how AI works, its limitations, and when AI is not always the right solution to a problem. Rather than stopping at mere glorification, AI’s opportunity in Indonesia can truly become a new source of optimism if the entire ecosystem is prepared seriously and strategically.
[1] Wadipalapa, R.P. et al. (2024) ‘An ambitious artificial intelligence policy in a decentralised governance system: Evidence from Indonesia’, Journal of Current Southeast Asian Affairs, 43(1), pp. 65–93. doi:10.1177/18681034231226393.
[2] Hidayat, T. (2026) Kesenjangan Digital: Meruntuhkan Ilusi Konektivitas, Menuntut Keadilan Akses, Birokrat Menulis. Available at: https://birokratmenulis.org/kesenjangan-digital-meruntuhkan-ilusi-konekt... (Accessed: 17 May 2026).
[3] Stryker, C. Apa itu human in the Loop (HITLl)? | IBM, Apa itu human-in-the-loop? Available at: https://www.ibm.com/id-id/think/topics/human-in-the-loop (Accessed: 27 April 2026).
[4] Kementerian Komunikasi dan Digital Republik Indonesia (2025) Berkas Publikasi: Indeks Masyarakat Digital Indonesia 2025, Pusat Pengembangan Ekosistem SDM Komunikasi dan Digital. Available at: https://bpsdm.komdigi.go.id/satker/pusbangesdmk/publikasi-indeks-masyara... (Accessed: 29 May 2026).
[5] Noer, H. and Pravitasari (2025) Menavigasi Adopsi Ai Dan ‘Super App’ di Sektor Publik Indonesia, GovInsider. Available at: https://govinsider.asia/indo-en/article/menavigasi-adopsi-ai-dan-super-a... (Accessed: 28 April 2026).
[6] Tuasikal, R. (2022) Ironi Indonesia 4.0 Dan Pemerintah yang hobi bikin Aplikasi Nirfaedah – project multatuli, Project Multatuli. Available at: https://projectmultatuli.org/ironi-indonesia-4-0-dan-pemerintah-yang-hob... (Accessed: 29 May 2026).
[7] Tucker, I. (2013) Evgeny Morozov: ‘we are abandoning all the checks and balances’ | technology | the guardian, Interview: Evgeny Morozov: ‘We are abandoning all the checks and balances’. Available at: https://www.theguardian.com/technology/2013/mar/09/evgeny-morozov-techno... (Accessed: 29 April 2026).
[8] Madan, R. and Ashok, M. (2023) ‘AI adoption and diffusion in Public Administration: A systematic literature review and future research agenda’, Government Information Quarterly, 40(1), p. 101774. doi:10.1016/j.giq.2022.101774.
[9] Putri, L.M. (2026) Polisi Tunda Pemeriksaan Freya JKT48 Terkait Kasus Penyalahgunaan Ai – Antara News Bengkulu. Available at: https://bengkulu.antaranews.com/berita/462119/polisi-tunda-pemeriksaan-f... (Accessed: 16 May 2026).
[10] Unggah Update Banjir Aceh, Afina Syifa Mengaku Jadi Target Buzzer (2026) Pos Kota. Available at: https://www.poskota.co.id/2026/05/29/unggah-update-banjir-aceh-afina-syi... (Accessed: 22 June 2026).
[11] Hutasoit, R.H. and Boen, H.S. (2025) Karya Tanpa Jiwa: Ai vs musisi Dan Hak cipta, Hukumonline. Available at: https://www.hukumonline.com/berita/a/karya-tanpa-jiwa–ai-vs-musisi-dan... (Accessed: 29 May 2026).
[12] Patrissia, R.U. (2026) Ai boom or bubble? Indonesia must choose readiness over hype – Antara News, ANTARA News. Available at: https://en.antaranews.com/news/411153/ai-boom-or-bubble-indonesia-must-c... (Accessed: 28 April 2026).
[13] AI adoption is boosting productivity, particularly among Indonesia’s Gen Z, yet skill gaps across generations remain a challenge (2026) PwC. Available at: https://www.pwc.com/id/en/media-centre/press-release/2026/english/hopes-... (Accessed: 29 April 2026).