With the rapid development of smart Internet of Things (IoT) devices, the security of private health data from unauthorized access is a significant concern. Although cloud-based analytics with smart devices have enhanced the health domain for automated health monitoring, the privacy concerns associated with the automated approach still persist. Traditionally, for automatic health data analysis, cloud-based analytics is adopted, which demands to transmit continuous data to the remote server for analysis. Therefore, health data analytics can be deployed at the IoT devices due to the specific level of computational capability of the contemporary IoT devices to transfer the cloud-based systems from the cloud to the edge devices to empower private health data security. In this chapter, we have selected the use-case of secure arrhythmia monitoring by utilizing Artificial Intelligence (AI) techniques to develop an efficient model that can be deployed at the edge devices for multi-class heartbeat classification to monitor electrocardiogram (ECG) of subjects for a prolonged time. For developing the automated model, we propose the random forest (RF) model by adopting state-of-the-art data analytics techniques and compare with four other Machine Learning (ML) methods. The guidance of ANSI/AAMI EC57:1998 standard is followed, and clinically rated MIT-BIH Arrhythmia dataset from PhysioNet is utilized for the secure multi-class classification system. We adopted diverse experimental setups to ensure the generalization capability of the proposed AI-aided RF model. The encouraging experimental results demonstrate the proposed model’s appropriateness for embedding it with the edge devices and achieving secure ECG monitoring without the need to share the data with remote servers for analytics and provide the edge devices with the processing ability.