Indoor positioning systems (IPS) continue to encounter significant challenges in achieving meter-level accuracy, particularly in large and intricate environments such as airports, hospitals, and industrial sites. Despite substantial advancements in technologies like Wi-Fi Fine Time Measurement (FTM), Ultra-Wideband (UWB), and 5G, their scalability and precision often suffer in dynamic, obstacle-laden settings. To enhance accuracy and adaptability, this study presents an innovative AI-driven sensor fusion technique that integrates machine learning models with various positioning technologies. This study achieves significant improvements in positioning accuracy through dynamic sensor input adjustments and real-time environmental awareness. Additionally, this study also includes results from a thorough evaluation of the system's effectiveness across diverse practical applications, while exploring its potential uses within the Internet of Things (IoT) framework.