ABSTRACT

The associability of artificial intelligence (AI) as a dual-use technology for nuclear material production (NMP) within the academic and practitioner communities remains widely neglected, and so a widening opportunity for AI to aid in illicit and covert non-peaceful applications exists. To address this emerging gap, this paper investigates the evolving and applicable uses of AI and finds broad evidence of its use to optimize performance, promote innovation, reduce costs, and enhance safety associated with the development and production of nuclear material. AI’s use in this arena will, thereby, facilitate broader accessibility of peaceful uses of nuclear science and technology, while at the same time cause concerns that said improvements can aid the illicit development of nuclear weapons. As such, this paper advocates for a three-dimensional solution to manage the evolving dual-use concern of AI that involves advancing states-centric monitoring and regulation, promoting intellectual exchange between the nonproliferation sector and the AI industry, and encouraging AI industrial contributions.

摘要

人工智能（AI）作为核材料生产（NMP）的双重用途技术，在学术界和从业者群体中的关联性仍然被广泛忽视，因此人工智能在帮助非法和隐蔽的非和平应用方面存在着越来越多的机会。为了解决这一新出现的差距，本文研究了人工智能的不断发展和适用用途，并找到了其用于优化性能、促进创新、降低成本和增强与核材料开发和生产相关的安全性的广泛证据。因此，人工智能在这一领域的使用将促进更广泛地和平利用核科学和技术，同时引起人们的担忧，即上述改进可能有助于核武器的非法发展。像这样，