The rapid growth of the Internet of Things (IoT) and mobile devices has generated massive volumes of data that require real-time processing for intelligent decision-making. Traditional cloud-based machine learning models face challenges such as high latency, bandwidth limitations, and privacy concerns. Edge Artificial Intelligence (Edge AI) has emerged as a promising paradigm that enables machine learning algorithms to run directly on IoT devices, smartphones, and edge servers, thereby reducing dependency on centralized cloud infrastructure. This paper explores the current state of Edge AI, highlighting advancements in lightweight deep learning models, hardware accelerators, and on-device training techniques. Real-world applications in smart cities, healthcare, autonomous vehicles, and industrial automation are discussed to demonstrate its transformative potential. Furthermore, the paper examines key challenges including limited computational resources, energy efficiency, model optimization, and security. By providing insights into both opportunities and barriers, this study emphasizes the critical role of Edge AI in shaping the next generation of intelligent and connected systems.