A NEW INSIDER VULNERABILITY ASSESSMENT MODEL

A new insider vulnerability assessment model is currently being developed at Lawrence Livermore National Laboratory (LLNL). The model focuses on timely detection of theft or diversion of nuclear material by a non-violent insider. This model will be part of an integrated vulnerability assessment package being developed jointly by LLNL and Sandia National Laboratory. The model will enable a safeguards analyst to complete a comprehen- sive and well-documented vulnerability analy- sis. Two major components are used in the model. The first consists of a description of the facility being analyzed, the safeguards compo- nents in use at the facility, the theft material, and the potential insider adversaries. The sec- ond component is a predefined set of strategy definitions, including specific methods of defeating the individual safeguards compo- nents. It also contains the logic and equations necessary to determine which strategies apply to the various insider adversaries and to com- pute the associated detection probabilities. The optimal theft scenario and the overall proba- bility of detection are determined for each type of insider defined. Use of predefined databases leads to consistency of results and ease of eval- uation. The reports generated are designed to aid in identifying weaknesses and suggesting potential upgrades. Re-evaluation is simplified by the structure of the model.