The swift expansion of Internet of Things (IoT) devices has resulted in heightened susceptibility to advanced cyber threats. Traditional centralized security models are no longer sufficient to handle the growing complexity, scale, and dynamic nature of modern IoT environments. This paper proposes a novel framework that integrates Artificial Intelligence (AI) with blockchain-based smart contracts to provide a scalable, decentralized, and real-time threat detection and response system for IoT networks. Our model leverages federated learning, enabling IoT devices to collaboratively detect security anomalies while preserving data privacy, and adaptive AI-driven smart contracts that autonomously adjust security policies in response to evolving threats. The proposed system also addresses critical challenges such as energy efficiency and scalability by incorporating lightweight consensus mechanisms and energy-aware decision-making for IoT devices. Furthermore, our framework facilitates cross-domain threat intelligence sharing through blockchain-based smart contracts, enabling multiple IoT networks to collaborate in detecting and mitigating threats across different industries, such as smart cities and healthcare IoT. This paper presents the theoretical underpinnings of the framework, detailing the architecture and expected performance improvements in detection accuracy, response time, and energy consumption. We hypothesize that the proposed system will outperform existing centralized solutions in terms of scalability, privacy protection, and real-time adaptability, offering a promising avenue for securing IoT networks in a decentralized manner.