Edge computing, which reduces latency and bandwidth usage by performing computations on data closer to where they are generated, is generating considerable interest due to the rapid growth of the Internet of Things (IoT) and real-time applications. However, this new architecture brings more security issues, such as cyber-attacks, unauthorized access, and data leakage. This architectural change improves performance, but it also increases the risk for data leakage, unauthorized access and cyberattacks. These distributed/architecturally decentralized scenarios are not something traditional cloud security models are suitable for. Federated learning (FL), a privacy-preserving setting for distributed learning, comes as a new solution which enables edge devices to collaboratively update their models without disclosing locally learned models. When integrated with AI, FL offers intelligent, flexible, and privacy-preserving security to edge ecosystems. This chapter investigates the interplay between edge computing, FL, and AI, providing a detailed analysis of possible future developments, risk mitigation strategies, and existing threats. This chapter studies the pioneering role of AI-supported federated systems in defending the future generation of edge networks through recent research and applied studies.