AI-generated urban health indicators: an empirical investigation from Egypt

Purpose Conventional urban health assessment frameworks frequently overlook the intricate, context-specific dynamics of the Global South. This study introduces an AI-assisted methodology for developing localised urban health indicators in Egypt, integrating generative artificial intelligence (GenAI) with expert validation to co-develop a context-sensitive Urban Health Index aligned with national and global agendas. Design/methodology/approach Three distinct large language models generated 48 indicators across the 8Ps-Framework. These indicators were validated by urban experts via a structured online survey. Psychometric analyses were conducted to confirm robust indicators. Findings Psychometric analyses confirmed 32 robust indicators (I-CVI = 0.78; CVR = 0.56 and ACP = 78%), demonstrating high internal consistency (Cronbach’s a = 0.948). Female and mid-career experts demonstrated greater confidence in AI-generated indicators, highlighting demographic influence on judgement. These findings underscore AI’s potential in co-creating adaptive, data-driven urban health metrics that complement expert judgement and expedite the index’s development. Research limitations/implications Despite expert sample size limitations and the specific focus on Egypt, this framework offers policymakers a scalable, cost-effective tool for rapid, locally relevant urban health assessments, supporting data-informed development and monitoring. Practical implications This study addresses the need for dynamic, context-specific metrics in resource-constrained contexts, offering a new approach to urban health assessment. Social implications The integration of AI with expert validation promotes inclusive, equity-focused health governance, fostering just and responsive urban settings, particularly in rapidly urbanising and resource-constrained Global South contexts. Originality/value This study presents a novel AI-assisted methodology for developing localised urban health indicators, providing a replicable model that addresses current methodological gaps and sets the stage for future AI-assisted urban health research.