1. This study seeks to identify factors to consider when formulating standards to manage the national security risks of using artificial intelligence (AI) for decision support systems. The goal is to start a dialogue on creating standards that will reduce the risk from use, misuse, and exploitation of AI, without impeding the United States’ technological development and competitive advantage.
译文：本研究旨在确定在制定有关将人工智能（AI）用于决策支持系统时进行国家安全风险管理方面的标准时需要考虑的因素。本研究的目的是就标准的制定展开对话，在不妨碍美国技术发展和竞争优势的前提下，减少人工智能的使用、滥用和利用的风险。
2. Public/private partnerships and education for users is critical to understanding AI technology and the associated threats, vulnerabilities, and risks.
译文：公共/私人伙伴关系和用户教育对于理解人工智能技术以及相关威胁、漏洞和风险至关重要。
3. To build trust, fairness, transparency, and accountability of AI to curtail error and misuse while ensuring functionality and securing against attacks, algorithms should undergo the “illities” test.
译文：为确保人工智能可信、公平、透明、负责，以减少错误和滥用，同时确保其功能性和免受攻击，算法应该进行“缺陷”测试。
4. Artificial intelligence (AI) likely is one of the most dramatic technological game changers of our time with the potential to transform human life from daily social interactions to how we conduct warfare. Specifically, AI will play a critical role in driving change in military, information, economic superiority, and the nature of security risks that will affect the ability of the United States to pursue its four pillars of national security.
译文：人工智能（AI）可能是我们这个时代最引人注目的技术游戏变革者之一，很可能将人类生活从日常的社会互动转变为我们如何进行战争。具体而言，人工智能将在推动军事、信息、经济优势和安全风险性质的变化方面发挥关键作用，这些变化对美国加强国家安全四大支柱的能力产生影响。
5. AI will have digital, physical, and political security implications, expanding existing threats, introducing new threats, and changing the character of threats and of war.
译文：人工智能将产生数字、物理和政治安全方面的影响，扩大现有的威胁，造成新的威胁，并改变威胁和战争的特征。
6. Among the key NDS objectives include defending the homeland; dissuading, preventing, or deterring state adversaries and non-state actors from acquiring, proliferating, or using weapons of mass destruction; and preventing terrorists from directing or supporting external operations against the United States homeland and our citizens, allies, and partners overseas.
译文：国防战略的主要目标包括保卫祖国；劝阻、防止或威慑国家对手和非国家行为者获取、扩散或使用大规模毁灭性武器的行为；防止恐怖分子指挥或支持针对美国本土以及美国的公民、盟友和海外伙伴的外部行动。
7. RESEARCH APPROACH: To understand governance, development, technology, algorithms, and potential standards, we conducted a literature review and interviewed stakeholders, including practitioners, developers, researchers, academics, and end-users from 19 organizations.
译文：研究方法：为了理解治理、开发、技术、算法和潜在的标准，我们进行了文献综述并访谈了利益相关者，包括来自19个组织的从业者、开发人员、研究人员、学者和最终用户。
8. We collaboratively developed a set of questions to frame our discussions with AI subject matter experts. These focused on understanding the technology and their thoughts on standards.
译文：我们共同提出了一组问题，作为我们与人工智能主题专家进行讨论的框架。这些问题专注于理解技术和专家们对标准的思考。
9. However, researchers suggest some AI systems are vulnerable to data leakage and unintended memorization, resulting in models that can reveal sensitive data to users.
译文：然而，研究人员认为，一些人工智能系统易受数据泄露和非预期记忆的影响，从而导致模型向用户泄露敏感数据。
10. We are still in the very early stages of understanding and discussing standards for AI. Consequently, we chose the following questions and thoughts on AI and standards to spur discussion to move towards a better understanding of how standards can apply to AI.
译文：我们仍然处于理解和讨论人工智能标准的最早期阶段。因此，我们选择了以下关于人工智能和标准的问题和想法来推动讨论，以便更好地理解标准如何适用于人工智能。