Asia Pacific Journal of Public Administration

Governing Artificial Intelligence in the Public Sector: Comparative Pathways in the Asia-Pacific and Beyond

Background and Motivation

Artificial intelligence (AI) is reshaping how governments make decisions, deliver services, and exercise authority. As machine learning, automated decision-making, and increasingly agentic systems move from experiment to infrastructure, they raise questions that are at once technical and profoundly administrative: how to preserve accountability, due process, and administrative justice; how to sustain public trust; and how to design the human, organisational, and regulatory arrangements through which AI is governed.

How governments answer these questions is not converging on a single model. The European Union has pursued comprehensive, risk-tiered regulation; the United States has favoured a more executive- and sector-driven approach; and across Asia a plurality of pathways is taking shape. China couples an ambitious developmental agenda with pervasive mechanisms of oversight and control; Japan balances AI promotion and safeguards through a flexible, risk-based approach; Korea pairs framework legislation with trust-building instruments such as generative-AI labelling and impact assessment; Taiwan emphasises principles, risk classification, and regulatory adaptation; and India is charting a democratic–developmental course of its own. These differences are not noise to be abstracted away. They are evidence of how political systems, administrative traditions, state capacity, and state–society relations shape the governance and adoption of AI.

This Special Issue takes that variation as its analytical engine. Rather than treating any region as exceptional or any model as universal, it adopts a comparative, context-sensitive stance: one that recognises how contextual differences shape administrative systems while insisting that the resulting knowledge remains comparable and transferable across jurisdictions. The aim is to move beyond both the assumption that AI governance produces similar outcomes everywhere and the opposite assumption that each setting is entirely unique and beyond comparison. Situating Asia’s diverse pathways alongside EU and US comparators allows the issue to contribute to a genuinely global understanding of how governments govern AI, and advancing public administration theory rather than documenting a single region.

The timing is deliberate. By the time this issue appears, the frontier of public-sector AI will have shifted further toward agentic systems and automation, human–machine collaboration, and the transparency and provenance of machine-generated content, including generative-AI labelling and watermarking. These developments sharpen enduring administrative concerns, including accountability, discretion, fairness, and control, while also generating new ones. Continuing the journal’s engagement with digital transformation and AI in government, this issue foregrounds the comparative governance of these emerging capabilities and welcomes rigorous, theory-informed, and practice-oriented scholarship, including cross-national and multi-level comparative studies and design-oriented contributions.

Aim and Scope

This Special Issue seeks contributions grounded in Asia-Pacific experience that advance global understanding of how governments govern and use AI. We welcome conceptual and empirical work, quantitative, qualitative, mixed-methods, and design-oriented that draws lessons from public-sector AI across jurisdictions and connects them to questions of broad, transferable relevance: how AI reshapes administrative discretion, the delivery and quality of public services, transparency, accountability, and the relationship between citizens and the state.

Single-jurisdiction studies are as welcome as cross-national comparisons, provided they speak to these wider questions and engage comparative or theoretical debates; comparative designs within Asia, and between Asia and regions such as the EU and the US, are especially valued for the leverage they give on how context shapes AI governance. We particularly encourage work on multi-level governance and on underrepresented jurisdictions, and work that develops or tests theory and makes a clear case for its significance to the journal's readership.

Indicative Topics

Submissions may address, but are not limited to, the following. Contributions are expected to connect their technological focus to questions of public governance, administration, and theory.

Comparative governance pathways

• Comparative AI governance pathways: how political systems, administrative traditions, and state capacity shape national approaches to regulating and governing AI; including framework legislation, risk classification, and the authorising relationship to national AI strategy and coordination bodies.
• Multi-level and cross-agency coordination: central–local division of authority, sectoral guidance, and the institutional design of AI strategy or coordination committees (chair seniority, expert and civil-society participation, cross-ministerial coordination).
• Global AI governance and AI for good governance: how robust global governance should be institutionalized for ensuring AI systems safe, equitable, and accountable or how different governments have been deployed AI technologies responsibly to build more resilient, trustworthy, and effective public institution.

AI and the work of public administration

• Automation, agentic AI, and administrative discretion: how automated decision-making and increasingly agentic systems reshape street-level discretion, professional judgement, accountability, and administrative justice in public organisations.
• AI, service delivery, and public value: the effects of AI on efficiency, effectiveness, quality, and equity of public services, and on citizens' experience of the state.
• Transparency, explainability, and contestability: algorithmic transparency and citizens' capacity to understand and challenge AI-informed decisions, with generative AI labelling requirements across jurisdictions as a comparative lens.
• Human–machine collaboration and administrative capacity: the design of human oversight (human-in-the-loop and human-on-the-loop), workforce reskilling, and organisational readiness for AI.

Accountability, trust, and public values

• Bias, equity, and administrative justice in public-sector AI.
• Accountability, oversight, and the legitimacy of the state: control, surveillance, and safeguarding public trust in government use of AI.
• Data governance, privacy, and public value: the stewardship of public data and the public-interest conditions for AI infrastructure.

Papers should make a strong case for their significance to the journal’s global readership; the editors especially value comparative studies and well-developed theory.