The rapid development of healthcare technology has enhanced clinical outcomes, personalized care, cost-effectiveness and resource utilization. However, this enrichment in the healthcare sector has numerous security issues, such as data breaches, phishing attacks, ransomware attacks, software vulnerabilities and insider threats (IT). Among those cyber attacks, the most redundant one is a data breach. The data breach can be an internal threat or an external threat. Most healthcare data breaches are affected by internal threats. The insider threat is unauthorized access to sensitive patient information with or without intention. Recently, the type of breach that affected patient’s records has been analyzed and determined by neural networks that can ensure the factors of insider activity and the evolution of insider threat in a data breach. This work presents an Agglomerative Nesting hierarchical clustering-based Attention neural network model to identify the indicators and evolution of insider threats to secure the health records of individuals affecting the healthcare industry. The proposed model was evaluated using a US healthcare data breach report regarding accuracy and detection rate. The performance of the proposed model is ensured by comparing it with different types of clustering-based neural network models.