This study investigates the application of generative AI (GenAI) tools (e.g., ChatGPT-4, DALL-E 3) in communicating climate resilience strategies across 15 global cities (2023–2025). Through a mixed-methods analysis of 92 GenAI-driven communication campaigns, 180 stakeholder interviews, and a public survey (n=4,200), it evaluates GenAI’s impact on public awareness of urban climate risks and engagement with adaptation measures. Results show GenAI enhances message personalization by 63% and increases voluntary participation in resilience projects by 38% compared to traditional communication. However, ethical risks—including misinformation propagation and algorithmic bias—are identified in 41% of campaigns. The research proposes a GenAI Ethics Framework for urban climate communication to guide responsible deployment.