Reshaping the Space of Ethics for Expert Work: Deliberative Approaches for Deploying Artificial Intelligence in Auditing
研究了领域专家与方法专家在审计中部署AI时的合作或冲突驱动因素，并基于协商民主理论提出通过多层协商重塑伦理空间以公平应对利益冲突的框架。
Abstract This paper investigates what drives collaboration or conflict between domain and method experts in deploying Artificial Intelligence (AI), and how these dynamics can be managed to fairly reshape the space of ethics guiding expert work. The emergence of sophisticated epistemic technologies, e.g., AI, has amplified the importance of method experts. Leveraging their knowledge of data science, mathematics, and coding, method experts have challenged the authority of domain experts, like external auditors. These challenges generate tensions extending beyond jurisdictional and operational control to profound ethical and moral concerns, since jurisdictional authority determines whose values are embedded in AI-driven decision-making. Empirical research presents divergent views on how domain and method experts navigate such tensions in AI deployments, with some studies suggesting that domain experts leverage their status and power to resist changes to ethical considerations. To reconcile these fragmented perspectives, we develop a conceptual framework grounded in the literature that frames expert work as a balance between individual craftsmanship and collective judgment. We offer two contributions. First, drawing on foundational literature on technology and expert work as well as empirical studies on AI and external auditing, we identify existing conditions and mechanisms that foster cooperation or conflict between different expert groups. Second, informed by Habermasian conceptions of deliberative democracy, we explicate how the space of ethics in deploying AI can be fairly reshaped through multi-layered deliberations that account for diverse stakeholder interests. Our framework provides a foundation for future empirical studies to deepen understanding of AI’s impact on expert work and ethics.