Adoption Frameworks for Artificial Intelligence in the Public Sector

Abstract

This systematic review aims to examine how the public sector may leverage artificial intelligence (AI) use through the exploration of AI adoption frameworks and how such frameworks can steer best practices of AI implementation and governance in the sector. Through inclusion and exclusion criteria, 30 articles were retrieved from academic databases, specifically Science Direct, Springer Link, and Wiley Online Library.The AI adoption frameworks are categorized into four groups: Regulatory frameworks, normative frameworks, applicative frameworks, and evaluative frameworks. Regulatory frameworks can provide standardizing and prescriptive guidelines to public sector organizations adopting AI technologies. Normative frameworks can strengthen the ethical and human rights aspects of AI adoption instead of devaluing human skills and eroding human agency. Applicative frameworks can help public sector organizations achieve positive and responsible outcomes for AI adoption. Alternatively, evaluative frameworks can spell improvements in the quality of public service delivery after identifying areas for improvement in an evaluation of AI systems.