Deliberating AI: Why AI in the Public Sector Requires Citizen Participation

Abstract

AI is a technology used across different societal domains due to its ability to increase operational efficiency by automating tasks and decision-making. AI is increasingly also applied in the public sector with the aim of making public administration more personalised, lean, and efficient. However, the technological advancement of AI also raises concerns regarding fairness, transparency, privacy, and human rights. Contrary to the perception of AI as a neutral tool, biases in its use and practices of surveillance suggest that it may have non-neutral impacts. Such impacts raise questions about AI's compatibility with democratic values. This thesis argues that if the use of AI in the public sector fails to uphold democratic values, its legitimacy is called into question based on democratic principles. The thesis investigates how AI in the public sector can impact the democratic values of equality, justice, and freedom. The investigation shows that AI should be understood as a political technology that risks disrespecting democratic values, raising questions about the legitimacy of its use in the public sector. The theory of deliberative democracy is employed to propose citizen participation as a way to address this. The thesis proposes the use of public deliberation as a means to determine how citizens would like to be governed by AI.