A socio-technical framework for AI trust in public administration

Purpose The purpose of this study is to examine the collective impact of transparency, explainable AI (XAI), participatory governance and digital literacy on public trust in AI within public administration, focusing on automated processing, AI-assisted decision-making and public service interactions. The study, from a socio-technical perspective, examines the mechanisms through which these elements foster trust and enhance citizen engagement with AI systems. Design/methodology/approach This research uses a systematic literature review, screening an initial pool of 272 publications to yield a final sample of 109 peer-reviewed articles for thematic analysis, synthesizing existing knowledge on AI trust in public administration. It proposes a novel integrated framework based on socio-technical systems theory. Findings Transparency, XAI, citizen engagement and digital literacy are identified as interconnected, context-dependent elements significant for public trust. Effective interaction between technical mechanisms (transparency, XAI) and socio-political structures (participatory governance, digital literacy) within a socio-technical system is posited to be essential for public trust. Research limitations/implications A key limitation is the framework’s conceptual nature. Its propositions require empirical validation through methods such as case studies, surveys or policy experiments to test its operational effectiveness and identify practical implementation challenges. Originality/value This study contributes a novel framework by theorizing the mutual constitution and reinforcement of technical and socio-political elements as a central mechanism for developing AI trust in public administration. It highlights the co-dependent nature of these elements.