The Role of Governance in Bridging AI Responsibility Gaps An interdisciplinary evaluation of emerging AI governance measures

The ubiquitous use of AI in a wide range of domain areas, including health, finance, social media, and many others, along with the well-publicised harms and concerns around the use of these systems, has generated questions around who is responsible, in the normative sense, for the outcomes of increasingly autonomous systems. Scholars within the interdisciplinary field of AI Ethics have noted that AI poses challenges to the attribution of moral and legal responsibility, due to the diminished knowledge and control of individual actors involved in bringing about system outcomes, and the existence of "many hands"- or a diffuse network of individuals and collectives who could potentially be responsible [6]. In my research, I draw from conceptual approaches in philosophy, examples from the history of technology, and domain-specific qualitative case studies to examine the extent to which AI presents new challenges to responsibility attribution. In addition, my research evaluates the effectiveness of emerging organisational and regulatory governance measures in meeting the challenges posed by apparent responsibility gaps.