Artificial Intelligence and the Objectified Production of Urban Space: Algorithmic Rationality and Urban Decision-Making

This conceptual article examines how artificial intelligence (AI) shapes urban governance. Rather than treating AI as a single, neutral tool, it breaks down the different forms AI can take: data infrastructures, statistical and machine-learning models, predictive systems, digital twins, and algorithmic decision-support tools. It examines when and how each gains influence in city decision-making. Drawing on Lefebvre’s theory of the social production of space and research on data-driven urbanism, algorithmic governmentality, platform urbanism, and AI urbanism, the article introduces the concept of the objectified production of urban space. This refers to the process of turning political and institutional decisions into data categories, indicators, model goals, predictions, and recommendations that then inform spatial decisions. This idea helps connect algorithmic mediation to specific outcomes in urban governance, such as how services, mobility, security, accessibility, investment, visibility, and risk are distributed. The article describes a five-stage process of objectification, identifies the key actors and types of authority involved, and explains when algorithmic mediation is more or less likely to create opacity and depoliticization. It argues that the pros and cons of urban AI should not be judged in general but should be considered from the perspectives of different stakeholders, public administrations, companies, investors, frontline workers, citizens, and affected communities, and based on technical, institutional, distributive, and democratic criteria. Short, documented examples and a discussion of algorithm registers, audits, impact assessments, human review, and appeal procedures show that these effects depend on the context rather than being inevitable. The article concludes by noting the limits of its analysis and suggesting a research program for comparative empirical testing. It also points out that the political aspects of urban AI are not only in the final decisions, but also in the technical details that shape how problems become data, indicators, thresholds, optimization goals, and recommended actions. Even though these outputs may seem rational, they still need critical examination. The key question is: who benefits from these decisions, and are they truly efficient or rational?