From the power broker to a(i) power broker

Abstract The rapid deployment of agentic artificial intelligence (AI) systems offers transformative opportunities for urban planning but introduces significant governance challenges. Drawing on historical lessons from Robert Moses and integrating insights from various indices and contemporary literature, this paper examines the risks of deploying AI agents with the potential to shape our cities without appropriate accountability mechanisms. The paper argues that urban AI must be governed through frameworks prioritizing transparency, ethics, public participation, and human-centric values to avoid reproducing inequities and to foster inclusive, adaptive city systems.