The Dawn of the AI Robots: Towards a New Framework of AI Robot Accountability
针对AI机器人应用中道德与问责缺失的问题，基于描述性和规范性伦理理论，构建了一个概念框架，分析道德主体位置与道德强度如何影响问责，并划分了非法、不道德、允许和超义务四类回应。
Abstract Business, management, and business ethics literature pay little attention to the topic of AI robots. The broad spectrum of potential ethical issues pertains to using driverless cars, AI robots in care homes, and in the military, such as Lethal Autonomous Weapon Systems. However, there is a scarcity of in-depth theoretical, methodological, or empirical studies that address these ethical issues, for instance, the impact of morality and where accountability resides in AI robots’ use. To address this dearth, this study offers a conceptual framework that interpretively develops the ethical implications of AI robot applications, drawing on descriptive and normative ethical theory. The new framework elaborates on how the locus of morality (human to AI agency) and moral intensity combine within context-specific AI robot applications, and how this might influence accountability thinking. Our theorization indicates that in situations of escalating AI agency and situational moral intensity, accountability is widely dispersed between actors and institutions. ‘Accountability clusters’ are outlined to illustrate interrelationships between the locus of morality, moral intensity, and accountability and how these invoke different categorical responses: (i) illegal, (ii) immoral, (iii) permissible, and (iv) supererogatory pertaining to using AI robots. These enable discussion of the ethical implications of using AI robots, and associated accountability challenges for a constellation of actors—from designer, individual/organizational users to the normative and regulative approaches of industrial/governmental bodies and intergovernmental regimes.