The perceptions of policymakers on the use of artificial intelligence in public administration

Abstract

As artificial intelligence (AI) becomes more deeply integrated into public administration, its influence on policymaking is drawing growing interest. In Malta, AI has been largely framed as a driver of economic progress, however, there is limited scholarly attention on how policymakers view its application in the policymaking process. This dissertation seeks to fill that gap by examining the perceptions of Maltese policymakers regarding the use of AI within public administration, with particular emphasis on its role in policymaking. This study adopts a qualitative approach, combining document analysis with semistructured interviews (N=9) to explore real lived experiences, perceptions, and institutional perspectives. Thematic analysis was employed to draw out key themes in the data, while triangulation was used to enhance the reliability, validity, and richness of the findings. The research is grounded in the Technology Acceptance Model (TAM) and supported by wider literature on digital governance and ethical AI. The findings suggest that while policymakers recognise the potential of AI to enhance efficiency and productivity, they remain cautious about issues related to transparency, accountability, excessive dependency on technology, and the diminishing role of human judgement. These results point to the need for AI governance strategies that are not only technical but also ethical and context aware. By shedding light on the unique challenges and opportunities of AI use in a small-state context, this study offers insight into the criteria necessary for responsible AI integration in public governance. As for future research, there is scope to examine the development of GPT-based models specifically designed locally and for public administration, along with the potential impact of Agentic AI in policymaking. In parallel, governance frameworks should evolve to ensure ethical integrity, contextual sensitivity, and institutional readiness in the adoption of such technologies.