A Comprehensive Method for Evaluating Safeguards Against the Insider Threat

Year
1989
Author(s)
T. Renis - Lawrence Livermore National Laboratory
R. Saleh - Lawrence Livermore National Laboratory
R.A. AI-Ayat - Lawrence Livermore National Laboratory
C.J. Patenaude - Lawrence Livermore National Laboratory
Abstract
We are developing a comprehensive tool for evaluating the effectiveness of safeguards and security systems against theft of special nuclear material by insiders. This tool, the Insider Evaluation module, is an integral part of ASSESS (Analytic System and Software for Evaluating Safeguards and Security) which is being developed jointly by Lawrence Livermore and Sandia National Laboratories. The Insider Evaluation module provides the user with a systematic, easy to use tool and an extensive database for analyzing the insider threat. This paper focuses on demonstrating the software, describing its input requirements and the various output options available to the user. An evaluation using this tool begins by defining the types of potential insider adversaries and specifying their levels of access and types of authority at the facility, as well as any procedures that may affect those theft strategies that take place across multiple facility layers. Users input to the module is facilitated by predefined databases of options from which the user can select those appropriate to their facility. The module identifies the optimal theft scenario and computes the corresponding probability of detection for each potential adversary type. Results are displayed in tables or graphs at varying levels of detail. For example, summary results present the overall probability of detection for each adversary. However, the user can focus, or \"zoom,\" on a particular result to gain further insights about the particular strategies or methods used by the adversaries. The most detailed results specify the best method an adversary could use to defeat each safeguards component at a path element to accomplish a particular strategy. Use of the predefined database of strategies and probabilities helps create consistent results and help document the modeling assumptions used.