Abstract: Against the backdrop of global climate change and accelerated urbanization, the frequency and severity of flood disasters have been increasing, posing significant threats to human life and properties. To explore a new paradigm of flood disaster prevention based on artificial intelligence (AI), we focus on three technical paths and application prospects: flood forecasting and early warning with decision support, intelligent watch keeping, and AI-assisted smart office applications. The results demonstrate that the AI-based rapid forecasting model, integrated with hydrodynamic simulations, achieves high accuracy with a coefficient of determination R² > 0.826 against simulated inundation depths, while improving computational efficiency by approximately 200~300 times compared to traditional models. Furthermore, the intelligent watch keeping system, which integrates meteorological monitoring, water-level sensing, digital human technology, and risk identification, can provide round-the-clock, efficient monitoring and response for urban flood management. In addition, the smart office system, powered by large language models, enables end-to-end automation in data processing and decision support. These findings offer valuable insights for advancing the intelligent development of flood disaster prevention systems.