This article proposes a research program that integrates recent advances in artificial intelligence with established theory and practice in drowning prevention, lifeguard operations, and aquatic risk management. Building on prior contributions that characterized the Instinctive Drowning Response and clarified the cognitive and perceptual demands placed on rescuers, I outline how intelligent systems can support earlier recognition, better prioritization, and more reliable intervention in aquatic environments. The paper introduces a modular architecture for AIenhanced water safety that spans perception, causal risk modeling, decision support, education and training, and system-level prevention. It presents practical evaluation criteria, interdisciplinary collaboration paths, and governance principles tailored to the realities of beaches, pools, waterparks, and open water. The agenda emphasizes human-centered design, ethical deployment, and translational research that connects laboratory methods with lifeguard stands, facility control rooms, and public health practice.