Conflict between states in the modern era takes place under the threat of nuclear weapons use. Preventing additional states, especially adversarial ones, from acquiring nuclear weapons is the goal of the United States Department of Defense’s Countering Weapons of Mass Destruction (C-WMD) program as defined in Joint Publication 3-40. This chapter analyzes the utility of machine learning in assessing specific indicators of nuclear proliferation based on feasibility and utility criteria. Nuclear proliferation indicators are developed and machine learning evaluation criteria designated and discussed. Implications for chemical and biological weapons are briefly discussed. A speculative look at far-future, true generalized artificial intelligence in the C-WMD fight is made, with a focus on determining new questions that could be answered by an advanced system. The results show that the most promising areas for machine learning in Counter-WMD are power grid analysis, imagery analysis to located hidden and protected sites, and communications metadata analysis to identify key players and their activity in proliferation networks. Far-future artificial intelligence may be able to track proliferator progress, anticipate nuclear decision points, and design new arms reduction frameworks.
CITATION STYLE
Exline, P. R. (2020). Machine learning in the countering weapons of mass destruction fight. In Advanced Sciences and Technologies for Security Applications (pp. 71–92). Springer. https://doi.org/10.1007/978-3-030-28342-1_5
Mendeley helps you to discover research relevant for your work.