Urban Algorithms and the Shape of the City: Recommender Systems in Municipal Infrastructure Planning

Featured in AI & Power Discourse Quarterly (ISSN 3080-9789), this study investigates the role of recommender systems in municipal infrastructure planning and urban governance. It discusses how algorithmic feedback loops influence city shaping, affecting planning decisions, resource allocation, and public participation, with implications for automated urbanism and smart city development. As artificial intelligence systems expand into urban governance, recommender algorithms—originally designed for consumer behavior prediction—are increasingly applied to infrastructure planning. This article explores how municipal governments in cities like Boston, Helsinki, and Singapore have begun integrating AI-based recommendation engines into public transportation, zoning, and emergency services allocation. The shift from deliberative urban design to data-driven selection introduces new logics of space: efficiency replaces debate, prediction substitutes consultation. By analyzing the structural biases of recommender systems, the article argues that algorithmic mediation in urban planning introduces a form of non-transparent determinism, where historical data silently dictate future development. This raises foundational questions about participation, access, and visibility in the algorithmic city. Rather than treating these tools as neutral amplifiers, the study frames them as active shapers of spatial power.