对偶(语法数字)
欺骗
计算机安全
信息技术
集合(抽象数据类型)
控制(管理)
困境
安全困境
计算机科学
业务
政治学
法学
人工智能
中国
艺术
哲学
文学类
认识论
程序设计语言
操作系统
作者
Jane Vaynman,Tristan A. Volpe
标识
DOI:10.1017/s0020818323000140
摘要
Abstract Almost all technology is dual use to some degree: it has both civilian and military applications. This feature creates a dilemma for cooperation. States can design arms control institutions to curtail costly competition over some military technology. But they also do not want to limit valuable civilian uses. How does the dual use nature of technology shape the prospects for cooperation? We argue that the duality of technology presents a challenge not by its very existence but rather through the ways it alters information constraints on the design of arms control institutions. We characterize variation in technology along two dual use dimensions: (1) the ease of distinguishing military from civilian uses; and (2) the degree of integration within military enterprises and the civilian economy. Distinguishability drives the level of monitoring needed to detect violations. When a weapon is indistinguishable from its civilian counterpart, states must improve detection though intelligence collection or intrusive inspections. Integration sharpens the costs of disclosing information to another state. For highly integrated technology, demonstrating compliance could expose information about other capabilities, increasing the security risks from espionage. Together, these dimensions generate expectations about the specific information problems states face as they try to devise agreements over various technologies. We introduce a new qualitative data set to assess both variables and their impact on cooperation across all modern armament technologies. The findings lend strong support for the theory. Efforts to control emerging technologies should consider how variation in the dual use attributes shapes this tension between detection and disclosure.
科研通智能强力驱动
Strongly Powered by AbleSci AI