Artificial Intelligence and the Transformation of Public Governance

Abstract

Artificial intelligence (AI) is already part of daily government work: chatbots, risk scores, forecasting models and tools for sorting applications faster than a human officer. This article looks at how that change is affecting public governance and how it’s affecting older public values like legality, fairness, accountability and equal access. The paper is conceptual and literature based. It draws on peer-reviewed works in public administration and regulatory works, as well as official texts from the OECD, UNESCO, the European Union, the Council of Europe, Canada, the United Kingdom, Singapore and the United States. The argument is clear, if not so much so. AI can improve efficiency, support more systematic evidence use and free staff from repetitive processing. The same systems can also reproduce bias and hide reasons for decisions, strain privacy and leave officials uncertain who is responsible when something goes wrong. Digital inequality makes these problems worse, because the people who depend on public services are most likely to be the least equipped to challenge an automated outcome. Scholarship has very well mapped applications and risks, but it is still somewhat less clear how administrative institutions absorb AI without strangling political responsibility. The article reviews uses in service delivery, policy analysis, administrative decisions and prediction and then compares various jurisdictional responses. Technical safeguards are not enough. What we need is institutional work: impact assessment, meaningful human review, public inventories of tools, procurement that doesn’t tie agencies into opaque systems, and investment in public sector skills. AI will keep spreading in government. And how much it improves public governance depends less on the models than on the legal and organisational arrangements built around them.