This article explores the use of AI-based adaptive protocols to optimize energy efficiency and reduce latency in IoT devices operating within fog-edge computing environments. As the number of IoT systems continues to grow, effective resource utilization has become essential for ensuring sustainable performance. The study compares AI-driven protocols with traditional approaches, highlighting key benefits in terms of energy efficiency, latency reduction, throughput improvement, and reliability. The results, derived from comprehensive simulations and performance assessments, demonstrate the adaptability of these protocols in response to varying operational conditions. The findings underscore the importance of incorporating AI applications into resource allocation strategies and provide a foundation for future innovations in IoT networks. The main conclusion emphasizes the continued exploration of AI-powered solutions to enhance IoT performance.