ABSTRACT

Automation is not new, but the possibilities for automation have been significantly expanded in recent years through advancements in artificial intelligence. Such technologies may drive some improvements, although they are not without risk and we lack a solid evidence base to suggest the implications of these changes. Framing AI supported automation as ‘technological outsourcing’, we draw on the well-established outsourcing literature to derive lessons about the possible implications of public sector automation and outline some principles that agencies can use to assist in their decision-making about whether to invest in automation of particular processes.

摘要

自动化并不新鲜，但近年来随着人工智能的进步，自动化的可能性得到了显着扩展。此类技术可能会推动一些改进，尽管它们并非没有风险，而且我们缺乏坚实的证据基础来表明这些变化的影响。将 AI 支持的自动化定义为“技术外包”，我们利用完善的外包文献来吸取关于公共部门自动化可能影响的经验教训，并概述一些机构可以用来帮助他们做出是否投资决策的原则特定过程的自动化。