This research focuses on the development of a novel AI-driven smart system for underwater navigation, communication, and environmental monitoring, aimed at enhancing diver safety and operational efficiency. The system integrates advanced technologies, including YOLOv8 for real-time object detection, MediaPipe for gesture recognition, and FLAN-T5 XL for natural language processing.Unlike existing smart diving solutions, which typically focus on individual features such as navigation or object detection, this system uniquely combines these components into a cohesive, multifunctional platform. This integration allows for hands-free operation through gesture and voice control, enabling real-time monitoring of both diver conditions and underwater surroundings.The system’s ability to recognize underwater objects, interpret diver gestures as commands, and provide immediate feedback on environmental conditions distinguishes it from traditional diving tools. This comprehensive approach makes the smart system an innovative solution for underwater exploration and safety, offering significant advancements in both technology and practical application. Keywords: smart system, convolutional neural network, computer vision, YOLOv8, MediaPipe.