During his electoral campaign, Donald Trump proclaimed in a speech on national security, immigration, and terrorism that the development of ‘a new screening test for the threats we face today’ was long overdue (Trump, 2016). Anybody with ‘hostile attitudes’ towards the US or ‘its principles,’ anybody supporting ‘bigotry and hatred,’ or unlikely to ‘flourish in our country’ should be screened ‘out’ (Trump, 2016). As the development of such procedures might take some time, he suggested that the processing of visas from a number of countries should be stopped for an unspecified time until the aforementioned procedures were in place (Trump, 2016). The measure was later realized as Executive Order 13769, the so-called ‘Muslim ban’ that targeted people from several Muslim majority countries – regardless of their Visa and passport status – from entering the US. Shortly after his inauguration, in January 2017, he renewed this pledge on Twitter, demanding ‘extreme vetting’ and ‘strong borders’ (Trump, 2017). Finally, in August 2017, leaked documents concerning an ‘Industry Day’ hosted by the US Immigration and Customs Enforcement’s Homeland Security Division painted a somewhat clearer picture of the proposed ‘screening test’ and ‘extreme vetting’ (Biddle and Woodman, 2017). The ICE had invited technology companies interested in the construction of such an ‘extreme vetting’ software and had presented its expectations in such a tool: An automated, centralized data-mining software capable of screening, vetting, and reviewing risk profiles in accordance with Executive Orders concerning ‘immigration and border protection security and interest’ (Biddle and Woodman, 2017). Furthermore, the would-be machine learning project was meant to ‘determine and evaluate an applicant’s probability of becoming a positively contributing member of society, as well as their ability to contribute to national interests’ and to predict ‘whether an applicant intends to commit criminal or terrorist acts after entering the United States’ (Biddle and Woodman, 2017; Harwell and Miroff, 2018). Information used for this purpose was meant to encompass all publicly available channels, including Social Media, telephone records, and geospatial databases etc., far beyond visa applications or criminal records (Biddle and Woodman, 2017).

中文翻译：

---

技术安全文化：简介

在竞选期间，唐纳德·特朗普在有关国家安全，移民和恐怖主义的演讲中宣布，早就应该开发“针对我们今天面临的威胁的新筛选测试”（特朗普，2016年）。任何对美国或其“原则”怀有“敌对态度”的人，支持“野心和仇恨”或不太可能“在我们国家蓬勃发展”的人都应该被“淘汰”（特朗普，2016年）。由于制定此类程序可能要花费一些时间，他建议应在未指定的时间内停止处理来自多个国家的签证，直到上述程序到位为止（Trump，2016年）。该措施后来作为13769号行政命令得以实现，所谓的“穆斯林禁令”是针对来自几个穆斯林多数国家（无论其签证和护照身分）的人进入美国。在就职后不久，2017年1月，他在Twitter上重申了这一承诺，要求进行``极端审查''和``严格边界''（特朗普，2017年）。最终，在2017年8月，有关美国移民和海关执法局国土安全部举办的``工业日''的泄露文件对拟议的``筛查测试''和``极端审查''进行了较为清晰的描述（Biddle和Woodman，2017年）。ICE邀请了对构建此类“极端审查”软件感兴趣的技术公司，并对该工具表示了期望：一种自动化的集中式数据挖掘软件，能够进行筛选，审查，并根据有关``移民与边境保护安全和利益''的行政命令审查风险状况（Biddle和Woodman，2017年）。此外，这个可能的机器学习项目旨在“确定和评估申请人成为社会积极贡献者的可能性及其为国家利益做出贡献的能力”，并预测“申请人是否打算犯罪”。或进入美国后的恐怖行为（Biddle和Woodman，2017； Harwell和Miroff，2018）。用于此目的的信息旨在涵盖所有公开可用的渠道，包括社交媒体，电话记录和地理空间数据库等，远远超出签证申请或犯罪记录的范围（Biddle和Woodman，2017年）。