Accountability and transparency in urban AI
![Monografia CIDOB nº 89](/sites/default/files/styles/max_width_150/public/2025-02/portada%20ingl%C3%A9s_1.jpg.webp?itok=9UA-c43N)
Many urban public administrations see the potential of implementing artificial intelligence (AI) but feel unprepared for how to do so responsibly. While responsible AI approaches and frameworks are increasingly popular, much of these focus on private or industrial actors. Though there is increasing attention to the role of responsible AI in the public sector, there is much less guidance for local governments specifically. This chapter aims to narrow the gap by presenting definitions of accountability and transparency that include both narrow technical and broader sociopolitical perspectives. To facilitate accountability and transparency in the context of implementing AI by urban public administrations, there are two deceptively simple yet fundamental guiding questions for the design phase: “Should AI be used?” and “How is AI to be used?” After reflecting on what these questions mean for urban practitioners, the chapter presents a summary of existing policy mechanisms which can be adapted to work towards these aims and offers some lessons learned from previous research.