Revolutionizing flood resilience: Human-centred AI for adaptive and regenerative urban design

Urban flooding poses escalating risks under climate change, requiring planning approaches that integrate technological precision with human and ecological values. This study introduces a Human-Centred Artificial Intelligence (HCAI) framework for adaptive and regenerative flood-resilient urban design. Unlike conventional AI approaches focused on optimisation, HCAI embeds ethical, participatory, and context-sensitive principles into land-use governance. The paper develops a four-pillar analytical framework – technical integration, institutional arrangements, stakeholder engagement, and land-use innovation – operationalised through a document-based comparative analysis of nine cities representing diverse flood typologies, governance regimes, and documented AI/digital flood-resilience activity. Documentary evidence suggests that AI tools such as Machine Learning, Deep Learning, and Graph Neural Networks are reported to support flood-risk prediction and response coordination in flood-risk applications, while their planning relevance depends on inclusive decision-making and regenerative land-use strategies. Cities with stronger institutional embedding of AI-enabled or digital tools within adaptive regulatory systems and participatory platforms (e.g., Rotterdam, Copenhagen, Bangkok) show clearer planning-policy pathways, whereas those relying on fragmented or project-based deployments (e.g., Venice, Sydney) show limited continuity. Findings suggest that technical capability alone is insufficient: planning relevance depends on governance maturity, ethical safeguards, civic legitimacy, and linkage to spatial planning instruments. By coupling algorithmic intelligence with participatory and regulatory innovation, the study advances a policy-oriented framework for embedding HCAI into spatial governance and identifying conditions that may support a shift from defensive flood control toward regenerative, socially just urbanism. The framework guides AI institutionalisation within land-use systems and contributes to debates on equitable, data-driven climate adaptation.