AI Decision Intelligence for Nuclear Material Accountability

This technical report discusses the application of trusted AI decision intelligence to enhance nuclear material control and accountability (MC&A). The document outlines the challenges faced by the nuclear industry in detecting human-centric threats, such as information theft and sabotage, and emphasizes the need for proactive, behavioral-science-driven risk models. Traditional security measures are described as often failing to identify early indicators of risk, necessitating a shift toward a more holistic approach. The report details how Decision Intelligence Platforms (DIPs) can integrate diverse data and provide explainable insights to support analysts in assessing insider risk. It presents a behavioral analytic modeling framework designed to monitor user behavior and detect anomalies, aiming to enhance security culture and reduce false positives. The report concludes that the integration of behavioral science with AI in decision intelligence platforms can facilitate proactive intervention in nuclear operations, improving resilience and effectiveness in MC&A.