Polibienestar is leading a project on the impact of AI on democratic legitimacy and public trust

The Research Institute of Social Welfare Policies (Polibienestar) at the University of Valencia is coordinating the project ‘Reasoning, Agency and Confidence in Human and AI-Enhanced Legitimacy’ (RACHAEL), which aims to analyse how artificial intelligence influences moral reasoning, ethical decision-making and citizens’ perceptions of legitimacy, justice and trust in democratic contexts. The initiative is led by Professor Jorge Garcés Ferrer and will run for three years.

The incorporation of artificial intelligence systems into public decision-making processes presents one of the democratic challenges of the coming years. RACHAEL is based on the premise that “these technologies can help to improve public innovation and support certain governance processes, but they can also affect people’s cognitive autonomy, the attribution of responsibility and trust in institutions”. In this vein, it seeks to generate scientific evidence on the conditions under which citizens accept or reject the involvement of AI in decisions with ethical, social and political implications.

To this end, the project combines experimental, quantitative and qualitative approaches. It comprises two sub-projects. The first, also known as RACHAEL, is led by Jorge Garcés and will conduct an experiment with a sample of 1,250 people residing in Spain to investigate whether knowing — or believing one knows — that a proposal has been drawn up by an expert or by an AI system alters perceptions of legitimacy, impartiality, fairness and trust. The dilemmas will be presented through fictional scenarios linked to areas such as the redistribution of resources, security, procedural justice, defence and rights. The design will allow researchers to observe not only the initial assessment of the decisions, but also how opinions change when the true authorship of the recommendations is revealed.

Meanwhile, the second sub-project, AImar, will analyse the personal, attitudinal and contextual factors that influence the acceptance of AI in democratic governance. The study will examine the role of trust in AI, institutional trust, pre-political values and cognitive delegation, understood as the tendency to transfer part of the deliberative effort to automated systems. To this end, a structured survey will be administered to the same sample of 1,250 participants, and a longitudinal qualitative follow-up will be carried out using in-depth interviews, aimed at understanding how citizens’ perceptions of AI evolve in political and ethical contexts.

Towards more transparent algorithmic governance

The team comprises experts in political science, psychology, public policy, marketing, behavioural economics and decision analysis. This interdisciplinary and socio-technical combination aims to link research on moral reasoning and trust with the design of more transparent, accountable and inclusive frameworks for algorithmic governance.

Furthermore, RACHAEL will develop a structured repository of real-world AI applications in the public sector, paying particular attention to cases relating to rights, resource allocation, access to services, institutional trust and citizen participation. Building on this work, the project will develop an analytical framework to evaluate algorithmic governance systems based on criteria such as transparency, explainability, accountability, public oversight, the capacity for rectification and institutional adaptation. The results will be validated with public officials, technical and legal experts, civil society organisations and citizen representatives through deliberative workshops.

The research team “hopes that RACHAEL’s findings will help guide the ethical and socially responsible integration of AI into democratic institutions”. According to Professor Garcés, “the project aims to provide useful evidence for public officials, social organisations and professionals working on the digital transformation of the public sector, helping to anticipate risks of mistrust, bias or loss of legitimacy when decisions are based on automated systems.” In this context, the University of Valencia is strengthening its role in research on democratic governance, public innovation and the social impacts of emerging technologies.

The initiative is funded with a budget of around 500,000 euros, as part of the 2025 call for research projects in the field of artificial intelligence issued by the Ministry of Science, Innovation and Universities and the State Research Agency.

About Polibienestar

Polibienestar is a public research institute at the University of Valencia, a leader in European projects of excellence, specialising in the analysis of public policy and the development of social innovations aimed at improving the functioning of European welfare systems. With a long track record in national and international projects, the institute carries out applied, interdisciplinary and socio-technical research in areas such as governance, sustainability, social innovation and citizen participation.

Read the news on the UV website: https://www.uv.es/uvweb/uv-noticias/es/noticias/polibienestar-lidera-proyecto-impacto-ia-legitimidad-democratica-confianza-ciudadana-1285973304159/Novetat.html?id=1286502146229&plantilla=UV_Noticies/Page/TPGDetaillNews

• Post Categories
• News