Date of Publication :February 2026

Abstract: This study describes the development and validation of an artificial intelligence (AI)-based risk assessment and tracking system. Drawing on machine learning (ML), the AI Risk Assessment App (AI-RAA) combines throughput-based risk assessment with a dashboard to track the progress of risk mitigation through the project life cycle. The AI-RRA was subjected to empirical field tests to validate its effectiveness based on 78 different types of real-world scenarios (e.g., construction, finance, information technology (IT), and health care) with an overall average accuracy of prediction of approximately 92.7% and an average consistent repeatability of 94%. More significantly, the AI-RRA reduced the risk assessment time by approximately 88% compared to traditional manual methods. We also evaluated how AI-RRA aligns with existing formal risk assessment methods, such as the National Institute of Standards and Technology (NIST) Artificial Intelligence Risk Management Framework (AI RMF) and the traditional risk mitigation hierarchy of controls. Our findings suggest that while the AI-RRA aids in accelerating hazard identification, it also standardizes the hazard output to produce a high level of trust and efficiency in risk workflows. Finally, discussions are provided regarding the implications, limitations, and future work on topics such as regulatory compliance and broader applications.

Reference :