Participatory AI Governance for Smart Cities and Sustainable Infrastructure

Abstract

Smart cities and sustainable infrastructure increasingly rely on artificial intelligence, connected sensors, digital twins, and automated decision systems. These technologies can improve energy efficiency, maintenance, mobility, and service delivery, but they also create risks related to surveillance, exclusion, cybersecurity, opacity, and concentrated institutional power. This review examines participatory AI governance as a missing element in the development of sustainable digital infrastructure. It builds on a systematic synthesis of 95 high-quality studies and governance frameworks covering the period from 2020 to mid-2025. The source analysis found that privacy, ethics, regulation, and accountability dominate the literature, while citizen participation, stakeholder engagement, and smart-city governance remain comparatively isolated. Technical and legal themes often form one cluster, ethical themes another, and participation a peripheral cluster with weak links to mainstream control design. The article argues that sustainable AI governance must combine engineering assurance with democratic involvement. It reviews participation across the AI lifecycle, from problem definition and data collection to deployment, monitoring, appeal, and retirement. It also applies the discussion to digital twins, structural health monitoring, energy management, transport, and public-service platforms. A practical participation-and-assurance model is proposed, linking stakeholder input to decision rights, technical controls, evidence, and public reporting. The review concludes that participation is not an optional communication exercise; it is a source of risk knowledge, legitimacy, and long-term system resilience.