AI Governance and Organizational Change in Developing States

The article begins by explaining that artificial intelligence and data science are increasingly transforming public administration globally. In developing countries, these transformations occur within bureaucratic systems that are fragmented, capacity-constrained, and politically layered. Indonesia is presented as an important case because of its large population, decentralized governance structure, and continuing digital government reforms.

The introduction emphasizes that Indonesia’s national initiatives in data integration and smart governance reflect an ambition to modernize state capacity through AI-driven systems. However, Indonesian public organizations continue to be shaped by hierarchical traditions and sectoral silos. This creates a complex institutional setting where technological tools interact with existing bureaucratic authority and professional norms.

The main problem identified is not simply whether AI can improve efficiency, prediction, or coordination. Instead, the article focuses on how AI systems are interpreted, filtered, and reshaped by organizational actors. In Indonesia, public agencies operate through layered authority arrangements, so AI tools can affect how decisions are made, justified, and controlled.

The article notes that AI implementation may reshape discretion among frontline officials, supervisors, and technical specialists. In policy areas such as social assistance targeting, regulatory enforcement, and digital service delivery, AI may redefine who has decision-making authority and on what basis. These changes have direct implications for accountability and public trust.

Existing scholarship on digital era governance has explained how data-intensive systems and intelligent centers restructure state functions. It shows that AI can enable centralized analytics, monitoring, prediction, and administrative holism. Other research has examined algorithmic bias, transparency, ethical safeguards, and the institutional nature of technology adoption.

However, the article argues that there is still limited understanding of how AI governance is mediated at the organizational level in developing-country bureaucracies. In particular, there is insufficient analysis of how AI reshapes power relations between central analytics units, line departments, and frontline officials. The role of technical elites, routines, and organizational resistance remains under-theorized.

The study positions Organizational Mediation Theory as a useful framework for addressing this gap. This theory views organizations as active filters that interpret, adapt, and reshape technologies rather than simply receiving them passively. Applying this perspective allows the article to move beyond technological determinism and examine how discretion, authority, power, and legitimacy are renegotiated internally.

The introduction concludes by stating the study’s objectives: to conceptualize AI governance in developing-country public sectors through Organizational Mediation Theory, analyze how AI reshapes bureaucratic discretion in Indonesian public organizations, examine redistribution of authority and power, and explore how legitimacy is constructed and contested when algorithmic systems support public decisions.

This study uses a qualitative research design with a theory-driven case study approach based exclusively on secondary data. The case focuses on AI and data-driven governance initiatives within Indonesian public sector organizations. A qualitative approach is considered appropriate because the study examines how AI reshapes discretion, authority, power relations, and legitimacy within complex bureaucratic structures. Organizational Mediation Theory serves as the analytical framework, treating public organizations as active filters that interpret, adapt, and reshape technological systems. The secondary-data design is suitable because AI governance reforms generate documentary evidence in official reports, policy statements, regulatory guidelines, academic literature, and public debates.

The study draws on national policy documents, ministerial regulations, institutional annual reports, parliamentary records, academic journal articles, NGO analyses, and credible media coverage related to AI and digital governance reforms in Indonesia. Sources were selected purposively, prioritizing documents that discussed AI implementation, data integration, digital transformation strategies, or organizational restructuring within public agencies. The analysis used deductive coding categories derived from Organizational Mediation Theory, including absorptive capacity, discretion redistribution, authority locus, data governance and integration, accountability and legitimacy, political embedding, service outcomes, and equity. Trustworthiness was strengthened through source triangulation, transparent coding, cross-referencing, and use of publicly available or properly cited materials.

The results show that AI and data-driven systems do not simply transform Indonesian public administration in a direct or automatic way. Instead, these systems become embedded within existing bureaucratic structures through organizational mediation. The article identifies absorptive capacity, discretion redistribution, and authority locus as three central mechanisms explaining how AI governance is shaped inside public organizations.

Uneven absorptive capacity is a major finding. National-level ministries often possess specialized digital transformation units with stronger technical literacy, training programs, and formal strategies for AI adoption. In contrast, sectoral departments and subnational agencies often have more limited capacity to interpret and operationalize algorithmic outputs. This creates dependency on central technical teams.

Because technical capacity is unevenly distributed, AI governance tends to concentrate epistemic authority in central units. These units become translators of AI-generated insights and gain influence over how data-driven tools are interpreted. As a result, AI does not diffuse evenly throughout public organizations but follows institutional capability gradients.

The study also finds that AI redistributes discretion within administrative hierarchies. Algorithmic systems are often introduced to standardize eligibility assessments, risk profiling, or compliance monitoring. These systems narrow the discretion of frontline officials in routine decisions, but they do not remove discretion altogether.

Instead, discretion shifts upward toward supervisory and technical actors who design, calibrate, and interpret algorithmic systems. Frontline officials may still exercise judgment in exceptional cases, but their decisions are increasingly shaped by algorithmically defined parameters. The article describes this as structured discretion compression combined with supervisory discretion expansion.

Another major finding concerns the reconfiguration of authority within public organizations. Although human officials formally retain final decision-making authority, algorithmic recommendations or risk scores create institutional pressure to follow system outputs. Technical experts who control data infrastructure and model parameters acquire informal authority that may exceed their formal rank.

This produces a hybrid authority structure. Formal bureaucratic hierarchy continues to exist, but it is increasingly combined with algorithmically anchored expertise. Technical elites become influential because they mediate data flows and system calibration, even when they do not formally occupy the highest administrative positions.

The study further shows that AI deployment is politically embedded. Policy narratives often present AI as a tool for transparency and efficiency, but political actors may selectively emphasize or downplay algorithmic outputs depending on strategic interests. This means that AI authority is not purely technical but is negotiated between administrative and political spheres.

The article argues that these findings refine Digital Era Governance theory. While Digital Era Governance suggests that AI and data-intensive systems create pressures toward intelligent centers and administrative holism, the Indonesian case shows that such outcomes depend on organizational learning, resource distribution, and institutional capacity. Centralization is therefore mediated rather than automatic.

The second part of the discussion examines broader governance consequences, especially data governance, accountability, legitimacy, service outcomes, and equity. Indonesian policy frameworks often emphasize data integration across ministries, but actual data sharing remains partial and shaped by sectoral mandates. Agencies may formally comply with integration goals while retaining control over important datasets.

Accountability is also re-routed rather than fully transformed. Official narratives often frame AI as improving transparency and evidence-based decision-making, but responsibility for outcomes remains embedded in existing bureaucratic hierarchies. Technical units gain operational influence over algorithmic systems, yet they may not face the same level of public scrutiny as formal decision-makers.

Finally, the article finds that AI governance can produce uneven service and equity outcomes. AI-driven targeting and regulatory systems may improve efficiency, but regions with stronger digital infrastructure and administrative capacity benefit more than weaker regions. In a decentralized state such as Indonesia, AI may reproduce or amplify existing institutional inequalities unless organizational capacity and intergovernmental coordination are also strengthened.

This study has examined how AI and data-driven governance reforms in Indonesia are shaped through organizational mediation rather than technological determinism. Beginning from the premise that Digital Era Governance pressures toward centralization and holism do not operate automatically, the analysis demonstrated that absorptive capacity, discretion redistribution, and authority locus function as core mediating mechanisms. Uneven absorptive capacity concentrates epistemic power within central technical units, while frontline discretion becomes structured within algorithmic parameters rather than eliminated. Authority is hybridized, combining formal bureaucratic hierarchy with emergent technical elites who control system calibration and data flows. These internal transformations subsequently shape broader governance consequences, including mediated administrative holism, re-routed accountability, constructed legitimacy narratives, and uneven service outcomes across regions. The findings collectively show that AI governance outcomes are institutionally filtered, politically embedded, and path-dependent. Rather than producing uniform modernization effects, AI systems interact with existing bureaucratic traditions and capacity disparities in Indonesia. Organizational mediation therefore provides the explanatory bridge linking technological adoption to governance consequences.

The study contributes to the field in three principal ways. First, it extends Digital Era Governance theory by embedding macro-level structural pressures within meso-level organizational dynamics, demonstrating that intelligent centres and holism are contingent upon institutional capacity and authority negotiations. Second, it advances Organizational Mediation Theory by applying it to AI governance in a developing-country context, thereby expanding its relevance beyond general technology adoption debates. Third, it enriches empirical scholarship on AI in public administration by shifting attention from normative algorithmic ethics toward intra-bureaucratic power reconfiguration and legitimacy construction. The analysis also challenges assumptions that AI inherently reduces discretion or enhances objectivity, showing instead that discretion is redistributed and legitimacy is narratively constructed. By situating Indonesia as a decentralized and capacity-uneven state, the study highlights how developing-country bureaucracies mediate digital transformation differently from highly institutionalized Western administrations. This contextualized contribution helps bridge global AI governance debates with Global South administrative realities. Overall, the article reframes AI governance as an organizationally negotiated process rather than a purely technical reform.

Future research should deepen comparative analysis across developing-country contexts to examine how variations in administrative tradition, decentralization, and fiscal capacity shape organizational mediation patterns. Cross-sector comparisons within Indonesia could further illuminate whether welfare, regulatory, and security domains exhibit distinct discretion and authority dynamics under AI adoption. Quantitative and mixed-method designs may complement secondary-data approaches by measuring changes in decision consistency, appeal rates, or regional equity outcomes. Longitudinal research would also be valuable to assess whether technical elites consolidate durable authority or whether hybrid arrangements stabilize over time. Additionally, future studies should investigate citizen perceptions of algorithmic governance to better understand how legitimacy narratives are socially received and contested. Greater attention to subnational governments is particularly important in decentralized states, where absorptive capacity gaps may widen governance inequalities. Finally, integrating political economy analysis with Organizational Mediation Theory could clarify how partisan competition and electoral incentives shape AI deployment strategies. Such research would further refine understanding of how AI governance evolves within institutionally diverse developing-country public sectors.

Aitken, A., Singh, S., & Otrisalova, S. (2024). Ageing and worker displacement. In Handbook on Labour Markets in Transition: Promoting Resilience in a World in Flux (pp. 389–423). https://doi.org/10.4337/9781839106958.00028

Alamsyah, A., & Aryfiyanto, H. (2025). Leveraging Generative AI for Public Service Innovation: A Path to Smart Government in Indonesia. Digital Government: Research and Practice, 6, 1–36. https://doi.org/10.1145/3761820

Alshehhi, O. (2025). The Role of Artificial Intelligence in Driving Change Management in the UAE Public Sector. International Journal of Technology and Systems. https://doi.org/10.47604/ijts.3479

Androniceanu, A. (2024). Generative artificial intelligence, present and perspectives in public administration. Administratie Si Management Public. https://doi.org/10.24818/amp/2024.43-06

Bingham, A. (2023). From Data Management to Actionable Findings: A Five-Phase Process of Qualitative Data Analysis. International Journal of Qualitative Methods, 22. https://doi.org/10.1177/16094069231183620

Bokhari, S., Park, S. Y., & Manzoor, S. (2025). Digital Government Transformation Through Artificial Intelligence: The Mediating Role of Stakeholder Trust and Participation. Digit., 5, 43. https://doi.org/10.3390/digital5030043

Bolia, L. M. G. (2025). Rethinking AI Readiness. Conference on Digital Government Research. https://doi.org/10.59490/dgo.2025.1069

Brizuela, V., Leslie, H. H., Sharma, J., Langer, A., & Tunçalp, Ö. (2019). Measuring quality of care for all women and newborns: how do we know if we are doing it right? A review of facility assessment tools. The Lancet Global Health, 7(5), e624–e632. https://doi.org/10.1016/S2214-109X(19)30033-6

Cheong, H.-I., Lyons, A., Houghton, R., & Majumdar, A. (2023). Secondary Qualitative Research Methodology Using Online Data within the Context of Social Sciences. International Journal of Qualitative Methods, 22. https://doi.org/10.1177/16094069231180160

Chiou, E. K., Holder, E., Dolgov, I., McDowell, K., Menthe, L., Roscoe, R. D., & Zaveri, S. (2020). Human, Ai, Robot Teaming and The Future of Work: Barriers and Opportunities for Advancement. Proceedings of the Human Factors and Ergonomics Society, 64(1), 62–66. https://doi.org/10.1177/1071181320641018

Creswell, A. J. (2018). Research Design: Pendekatan Kualitatif, Kuantitatif, dan Mixed. Pustaka Pelajar.

Dunleavy, P., & Margetts, H. (2023). Data science, artificial intelligence and the third wave of digital era governance. Public Policy and Administration, 40, 185–214. https://doi.org/10.1177/09520767231198737

Elbardan, H., & Kholeif, A. (2017). An Interpretive Approach for Data Collection and Analysis. 111–165. https://doi.org/10.1007/978-3-319-54990-3_5

Engstrom, D. F., & Ho, D. E. (2020). Algorithmic accountability in the administrative state. Yale Journal on Regulation, 37(3), 800–854.

Esmailzadeh, Y., & Motaghi, E. (2024). International Terrorism and Social Threats of Artificial Intelligence. Journal of Globalization Studies, 15(1), 168–179. https://doi.org/10.30884/jogs/2024.01.09

Hakimi, M., Zarinkhail, S., & Sahnosh, F. A. (2025). Artificial Intelligence and Legal Reform in Developing Countries: Advancing Ethical, Rights-Based, and Accountable Digital Governance. Jurnal Ilmiah Telsinas Elektro, Sipil Dan Teknik Informasi. https://doi.org/10.38043/telsinas.v8i2.6934

Karlekar, P. V, Choudhary, S., Deshmukh, A., & Banote, H. (2023). Innovations in Urban Automation: Robotic Arm Integration in Smart City Environments. 2023 International Conference on Sustainable Emerging Innovations in Engineering and Technology, ICSEIET 2023, 314–319. https://doi.org/10.1109/ICSEIET58677.2023.10303638

Kattof, A. L. A. A. (2025). The role of artificial intelligence in administive organization and combating corruption. The International and Political Journal. https://doi.org/10.31272/ipj.i62.423

Li, Y., Fan, Y., & Nie, L. (2023). Making governance agile: Exploring the role of artificial intelligence in China’s local governance. Public Policy and Administration, 40, 276–301. https://doi.org/10.1177/09520767231188229

Lochmiller, C. (2021). Conducting Thematic Analysis with Qualitative Data. The Qualitative Report. https://doi.org/10.46743/2160-3715/2021.5008

Mainardi, I. (2024). Change management: artificial intelligence (AI) at the service of public administrations. AI & SOCIETY, 40, 3953–3981. https://doi.org/10.1007/s00146-024-02136-2

Margetts, H. (2022). Rethinking AI for Good Governance. Daedalus, 151, 360–371. https://doi.org/10.1162/daed_a_01922

Misra, H. (2021). Amalsad cooperative ltd: total cost of ownership for digital services. Emerald Emerging Markets Case Studies, 11(4), 1–36. https://doi.org/10.1108/EEMCS-12-2020-0423

Morgan, H. (2022). Conducting a Qualitative Document Analysis. The Qualitative Report. https://doi.org/10.46743/2160-3715/2022.5044

Mukhlis, M. M., Maskun, Tajuddin, M. S., Andriani, D., Muchtasar, R., & Masum, A. (2025). Regional Government Autonomy in Indonesia: The Ambiguity of the Federalism of Republic Model. Malaysian Journal of Syariah and Law, 13(1). https://doi.org/https://doi.org/10.33102/mjsl.vol13no1.760

Mutale, W., Vardoy-Mutale, A.-T., Kachemba, A., Mukendi, R., Clarke, K., & Mulenga, D. (2017). Leadership and Management Training as a Catalyst to Health System Strengthening in Low-Income Settings: Evidence From Implementation of the Zambia Management and Leadership Course for District Health Managers in Zambia. Plos One, 12(7), e0174536. https://doi.org/10.1371/journal.pone.0174536

Napang, M., Nur, S. S., Bachril, S. N., & Al Mukarramah, N. H. (2021). Detrimental impact of Indonesian food estate policy: Conflict of norms, destruction of protected forest, and its implication to the climate change. IOP Conference Series: Earth and Environmental Science, 824(1). https://doi.org/10.1088/1755-1315/824/1/012097

Omonov, M., & Ahn, Y. (2025). Towards Smart Public Administration: A TOE-Based Empirical Study of AI Chatbot Adoption in a Transitioning Government Context. Administrative Sciences. https://doi.org/10.3390/admsci15080324

Purwanto, E. A. (2019). Kebijakan Publik Yang Agile Dan Inovatif Dalam Memenangkan Persaingan Di Era Vuca (Volatile, Uncertain, Complex and Ambiguous). Molecules, 9(1), 148–162. http://jurnal.globalhealthsciencegroup.com/index.php/JPPP/article/download/83/65%0Ahttp://www.embase.com/search/results?subaction=viewrecord&from=export&id=L603546864%5Cnhttp://dx.doi.org/10.1155/2015/420723%0Ahttp://link.springer.com/10.1007/978-3-319-76

Ruggiano, N., & Perry, T. (2017). Conducting secondary analysis of qualitative data: Should we, can we, and how? Qualitative Social Work, 18, 81–97. https://doi.org/10.1177/1473325017700701

Saputra, W. N., Ema, I., Sari, I. N., & Ramadhani, Q. T. (2024). Transformasi Birokrasi Digital dalam Pelayanan Publik: Studi Kasus Penerapan E-Government. Jurnal Sosial, Ekonomi, Dan Humaniora (SOSIERA), 3(2), 82–94. http://dx.doi.org/10.21776/jki.2024.03.1.23.

Schafheitle, S., & Meijerink, J. (2024). Trust and control amidst intelligent technology at work: What we can learn from online labor platforms. In A Research Agenda for HR Analytics (pp. 167–184). https://doi.org/10.4337/9781035301096.00017

Schlunegger, M. C., Zumstein-Shaha, M., & Palm, R. (2024). Methodologic and Data-Analysis Triangulation in Case Studies: A Scoping Review. Western Journal of Nursing Research, 46, 611–622. https://doi.org/10.1177/01939459241263011

Sharmin, S., & Chowdhury, R. H. (2025). Digital Transformation in Governance: The Impact of e-governance on Public Administration and Transparency. Journal of Computer Science and Technology Studies. https://doi.org/10.32996/jcsts.2025.7.1.27

Sigfrids, A., Nieminen, M., Leikas, J., & Pikkuaho, P. (2022). How Should Public Administrations Foster the Ethical Development and Use of Artificial Intelligence? A Review of Proposals for Developing Governance of AI. 4. https://doi.org/10.3389/fhumd.2022.858108

Tavberidze, N. (2025). Artificial Intelligence as a Development Tool for Digital Governance. Vectors of Social Sciences. https://doi.org/10.51895/vss9/tavberidze

Vatamanu, A.-F., & Tofan, M. (2025). Integrating Artificial Intelligence into Public Administration: Challenges and Vulnerabilities. Administrative Sciences. https://doi.org/10.3390/admsci15040149