Introduction Artificial intelligence (AI) is often described as a transformative digital technology, comparable to the internet or mobile connectivity. For governments, it offers opportunities to improve decision-making, deliver faster and fairer public services, manage resources more efficiently, and improve people's lives. Yet AI is developing and being adopted so rapidly that regulatory frameworks and safeguards are struggling to keep pace. Developing countries face complex challenges, including urbanization, extreme weather events, fiscal constraints, gaps in public service delivery, demographic change, and rising public expectations. AI could help governments respond to these pressures, but realizing its benefits requires more than access to technology. The central question is how governments can adopt AI responsibly, build the institutional foundations needed for its effective use, and deliver public value. The Development Challenge Is Institutional The quality of public institutions shapes development outcomes. Government effectiveness influences how efficiently revenues are collected, how transparently resources are managed, and how reliably citizens can access services. AI can analyze large volumes of information, identify patterns, forecast demand, automate routine tasks, retrieve institutional knowledge, and support engagement across language and geographic barriers. However, AI is not a remedy for institutional weaknesses. When deployed with fragmented or low-quality data, unclear mandates, weak safeguards, or poorly designed processes, AI may create new problems rather than resolve existing ones. Citizens need confidence that AI-supported services and decisions are fair, secure, explainable, and subject to review and redress. Governments therefore need to strengthen data quality and governance, interoperability, cybersecurity, public sector skills, and institutional accountability alongside AI adoption. Digital public infrastructure—including reliable electricity and broadband, digital identity, payment platforms, trusted registries, interoperable databases, and secure data-sharing mechanisms—provides important foundations for AI-enabled services. Readiness also depends on financing, computing infrastructure, governance frameworks, and capable public agencies. For governments with limited resources, however, these foundations need not be built all at once. The more practical question is how to sequence AI adoption while protecting inclusion, accountability, transparency, and public trust. Start with the Public Problem, Not the Technology Governments should begin with a clearly defined public problem or institutional constraint rather than a particular model, technology, or service provider. AI should be considered where it could help address a genuine service delivery or operational challenge. The appropriate approach will depend on the task: generative AI can support language-based work and information retrieval, while analytical, visual, or other forms of AI may be better suited to different problems. Once a potential benefit is identified, public agencies should examine the relevant processes, data, legal mandates, institutional ownership, risks, and user needs. They can then establish the safeguards and capabilities needed for a specific use case, test a bounded, relatively low-risk application with human oversight, and assess whether it creates public value. Expansion should follow only when evidence and institutional readiness support it. Because generative AI is already being tested and used across government institutions, cross-agency information sharing and peer learning should begin early. Communities of practice can spread skills, identify promising applications, and help teams communicate limitations and risks. Public servants need room to experiment within clear guardrails, with managers able to respond to problems openly and promptly. Maintaining public confidence is essential: trust can be lost quickly and is difficult to rebuild. Practical Entry Points for Capacity-Constrained Governments Countries with limited digital infrastructure or institutional capacity can begin with focused applications that address service delivery bottlenecks without making high-stakes decisions. Internal government work involving large amounts of information, repetitive processes, and slow retrieval of official knowledge may offer manageable starting points. Giving trained public employees access to appropriate AI tools can also help them understand the technology's capabilities and limitations while supporting their work. A subsequent step may be a citizen-facing service assistant that helps people find accurate information on eligibility, procedures, documents, fees, deadlines, and responsible agencies. Rather than connecting immediately to every government database, a pilot could cover a small number of frequently used services from one ministry or portal. It could draw answers from approved official information, link users to government forms and webpages, and refer uncertain or complex cases to a public official. An existing website or widely used messaging platform could serve as the initial channel. Uzbekistan's MyGov AI illustrates this information-and-navigation approach. The public platform describes assistance with government services and legal information, including service navigation, document preparation, and voice-enabled support. Such applications can provide a useful layer over existing services without requiring an immediate rebuild of the underlying delivery systems. Lessons from the People's Republic of China and Across Asia The PRC offers AI examples moving beyond experimentation into government and sector operations. Applications discussed during the AIxADB 2026 workshop included agricultural compliance and industry analysis, gig employment matching and risk control, economic monitoring, intelligent customer service, and digital knowledge retrieval. Together, these applications illustrate the importance of connecting AI with sector knowledge, government data, operational processes, and clear institutional responsibilities rather than treating it as a standalone tool. The PRC's open-source model ecosystem and extensive data resources also offer experiences for other economies to examine and adapt to their own circumstances. India offers a complementary lesson: AI adoption can build on broader digital transformation. Investments in digital identity, payments, public service platforms, and data ecosystems have established foundations for digital government services. The IndiaAI Mission, launched in 2024, seeks to expand computing capacity, access to datasets, domestic AI capabilities, skills, and responsible adoption, including access beyond major technology firms. India's experience highlights why AI should strengthen, not substitute for, reliable data systems and institutional foundations. Other economies in Asia and the Pacific offer related perspectives. Singapore highlights digital governance and whole-of-government coordination; the Republic of Korea illustrates the contribution of digital technologies to citizen-centered services; and Japan points to AI's potential to help address health care needs associated with population aging. These experiences are not models to replicate wholesale. Governments need to adapt relevant elements to their own mandates, capacities, and development priorities. Priority Opportunities and Safeguards With appropriate foundations, AI can support government effectiveness through knowledge systems, administrative automation, multilingual service assistants, and evidence-informed decision-making. In public financial management, it may assist with budgeting, expenditure monitoring, procurement oversight, auditing, fiscal risk management, and tax and customs compliance. In resilience and disaster response, it can combine satellite, weather, geospatial, and administrative data to support early warning, vulnerability mapping, and response efforts. Municipal services may benefit from AI applications in traffic management, maintenance, permitting, and waste collection. AI also has potential applications in health and education, including diagnostics, disease surveillance, telemedicine, teacher support, and school administration. For businesses, it can assist with credit assessment, market intelligence, quality control, logistics, energy efficiency, and supply chain decisions. These opportunities must be assessed against actual needs and implementation capacity rather than treated as guaranteed outcomes. Governments must also manage risks related to exclusion, cybersecurity vulnerabilities, dependence on external providers, privacy failures, and poorly governed decisions. Priority Opportunities for Developing CountriesAl can help governments deliver better services, stronger institutions, and more inclusive, resilient, and prosperous societies. Source: Authors. Supporting Responsible AI Adoption Development partners, such as the Asian Development Bank (ADB), can help developing countries translate AI's potential into development outcomes through support for policy reform, public institutions, infrastructure, financing, and implementation. Support may include advice on national strategies, readiness assessments, data governance and cybersecurity reforms, digital government modernization, digital public infrastructure, and public sector capacity development. Technical assistance can support diagnostics, skills development, and pilot initiatives. Policy-based lending can advance reforms in digital government, cybersecurity, and data governance. Investment projects can finance data and digital public infrastructure and AI-enabled services, while results-based lending can link financing to measurable improvements in institutions and service delivery. Across these instruments, the priority should remain the public problem to be solved and the capabilities needed for responsible implementation. AI as a Driver of Economic Growth AI offers developing countries opportunities to accelerate economic transformation and overcome some traditional barriers to development. The ADB simulations suggest that AI could increase gross domestic product growth by 0.6–2.1 percentage points in advanced economies in Asia and the Pacific by 2030, depending on the pace of AI adoption. For developing economies in the region, the estimated gains are more modest, at 0.2–1.8 percentage points. These projections illustrate AI's potential economic benefits while underscoring the importance of strengthening the foundations needed to realize them. Countries that combine responsible AI governance with targeted investments in digital infrastructure, workforce skills, research and innovation, and institutional reform will be better positioned to use AI to support more inclusive economic growth. Asia and the Pacific's young workforce is an important asset. Although automation raises legitimate concerns about job displacement, AI also offers opportunities to improve productivity, complement workers' skills, and create new forms of employment. With appropriate investments in education, training, and workforce development, AI could help turn the region's demographic potential into broader economic and social gains. This article draws on insights and observations informed by the AIxADB 2026 Integrated Governance and Seamless Services workshop held in Shanghai, People's Republic of China, on 8–11 September 2026. Ask the Experts Çiğdem Akın Director, Public Sector Management and Governance Sector Office, Sectors Group, Asian Development Bank Çiğdem Akın is director of ADB’s Public Sector Management and Governance Sector Office. Her areas of expertise include public financial management, fiscal policy, state-owned enterprise reforms, digital technology in public finance, and local governance. Prior to joining ADB, she worked at the Johns Hopkins University School of Advanced International Studies, George Washington University, International Monetary Fund, and Asian Development Bank Institute. She holds a PhD in economics from George Washington University and a master’s degree in international development from the International University of Japan. Arndt Husar Principal Operations Specialist (Operations Coordination Unit), South Asia Department, Asian Development Bank Arndt Husar leads knowledge management, capacity development, communications, and digital transformation for ADB’s South Asia Department. Previously, as senior public management specialist in the Digital Technology for Development Division, he provided technical assistance and supported knowledge sharing and capacity development in data management, technology risks, technology start-up ecosystems, and digital transformation strategies. Before joining ADB, he worked with the United Nations Development Programme at the national, regional, and global levels. Follow Arndt Husar on Pingpanya (Ping) Phommilath Public Sector Specialist, Public Sector Management and Governance Sector Office, Sectors Group, ADB Pingpanya (Ping) Phommilath is a public sector specialist at ADB, working on governance reform, digital transformation, and public sector modernization. His work focuses on institutional strengthening, public service delivery, and the use of digital technologies and artificial intelligence. Prior to joining ADB, he worked with United Nations agencies and the World Economic Forum on digital government initiatives. He holds a Master of Public Policy from Tsinghua University, a Master of Science from the University of Geneva, and a bachelor’s degree from Fudan University. Asian Development Bank (ADB) The Asian Development Bank is a leading multilateral development bank supporting sustainable, inclusive, and resilient growth across Asia and the Pacific. Working with its members and partners to solve complex challenges together, ADB harnesses innovative financial tools and strategic partnerships to transform lives, build quality infrastructure, and safeguard our planet. Founded in 1966, ADB is owned by 69 members—49 from the region. Follow Asian Development Bank (ADB) on Leave your question or comment in the section below: View the discussion thread.