Adrian Gavornik (Kempelen Institute Intelligent Technologies) and Juraj Podrouzek (Kempelen Institute Intelligent Technologies) have posted “Towards Moral Sensitivity in AI Practice” on SSRN. Here is the abstract:
This position paper introduces the concept of moral sensitivity into the context of AI ethics as a potential bridge between abstract ethical principles and their practical implementation. Besides this, we argue that fostering moral sensitivity among AI practitioners can help address active responsibility gaps in AI development. Building upon existing literature and empirical experience, we propose and outline several key components of moral sensitivity specific to AI practice. We present a dual research agenda for studying moral sensitivity in AI practice: a top-down approach analyzing how existing frameworks and tools foster different components of moral sensitivity and a bottom-up empirical investigation of how moral sensitivity virtue manifests in AI practitioners during ethical interventions. The paper concludes by discussing key challenges in studying moral sensitivity in AI practice, including methodological difficulties in empirical assessment, sustainability of measuring moral sensitivity over time, appropriate levels of abstraction, and the complexity of team dynamics in AI development.