The accelerated technological transformation that public administrations are undergoing presents new opportunities to connect with citizens, but it also highlights the limitations of traditional models of citizen participation, which are often symbolic and have little impact on public decision-making. This study proposes evolving towards participatory procedures based on citizen science, utilising open data and hybrid intelligence that combines collective and artificial intelligence (AI) capabilities to generate public value.
The aim of this research is to define and validate a framework for citizen co-governance, supported by digital platforms and learning algorithms, that can automatically transform the data generated by citizens into concrete public actions. To this end, we conduct a systematic analysis of the literature on participation and citizen science, from the classic vision of Arnstein (1969) to the current proposals of Fung (2006), OECD (2018) and Ataman (2025), which integrate inclusion and democratic legitimacy, identifying a gap in the operational mechanisms that translate citizen participation into tangible decisions. This analysis is complemented by the study of the CitizenBack pilot, developed in the Urban Sandbox of Valencia, which acts as a transversal government infrastructure and citizen digital twin. The pilot has been used to empirically evaluate data flows, citizen interactions, and the response mechanisms of AI algorithms, generating qualitative data, such as user perceptions, and quantitative data, including usage metrics, participation patterns and assessments of public management. Additionally, it is complemented by interviews with public managers to gain a deeper understanding of the institutional challenges involved in adopting such systems.
The results propose a new organisational model that automates public collection, analysis, prioritisation and response in real-time. Although there are challenges, such as the digital divide, algorithmic ethics, and change management within institutions, the study provides an operational framework for achieving data-driven, people-centred and AI-driven public co-management, thereby strengthening citizen democracy.
