The growing deployment of generative artificial intelligence (GenAI) in urban governance has intensified global debates on trustworthy AI. However, existing research remains largely technology-centric and offers limited empirical insight into how citizens perceive trustworthy AI and how these perceptions are translated into policy legitimacy. Addressing this gap, this study aims to examine how and for whom trustworthy AI generates legitimacy by analysing the mechanisms through which perceived GenAI trustworthiness is associated with citizens’ trust in urban AI policy.
Drawing on public value theory, institutional trust scholarship and sociotechnical governance perspectives, this study proposes a mediation model in which perceived institutional transparency and perceived public value shape the relationship between perceived GenAI trustworthiness and citizen trust in AI-enabled urban policy. A cross-sectional survey (n = 527) was conducted among residents of Malaysia’s Klang Valley metropolitan region using stratified quota sampling. The data were analysed using structural equation modelling.
The findings indicate that technical trustworthiness alone is insufficient to foster citizen trust in GenAI policy decisions. Perceived public value emerged as a significant mediating mechanism, whereas institutional transparency did not exhibit a significant mediating effect. These results suggest that policy legitimacy is shaped less by procedural transparency in isolation and more by whether GenAI is perceived to generate tangible and socially meaningful outcomes.
The study relies on cross-sectional, self-reported data from a single metropolitan region. Accordingly, the findings should be interpreted as associational rather than causal. Future research could use longitudinal, comparative or qualitative approaches to examine how citizens’ interpretations of AI governance, transparency and public value evolve across contexts.
The findings highlight the need to reframe trustworthy AI as a governance challenge rather than a purely technological achievement. For policymakers and urban administrators, prioritising public value creation, interpretable and citizen-oriented transparency and institutional accountability is critical to sustaining democratic legitimacy for GenAI-enabled policies.
This study advances research on AI governance by conceptualising trustworthy AI as a relational and institutionally embedded construct, shifting the debate beyond technical compliance toward the governance practices and public value considerations that shape AI legitimacy in urban contexts. By empirically demonstrating the differential roles of public value and transparency, the study provides a nuanced and context-sensitive contribution to understanding trust formation in AI-enabled governance.
