No One Decided Part III: Accountability Collapse in Algorithmically Mediated Organisations

This paper examines the phenomenon of accountability collapse in algorithmically mediated organisational decision-making. Building on earlier analyses of disappearing human presence and filtered access to care, it argues that contemporary algorithmic systems increasingly enable organisations to produce consequential outcomes without any identifiable decision-maker being answerable for them. Responsibility is not denied, but dispersed across systems, procedures, and roles in ways that render moral and legal accountability structurally elusive. Drawing on organisational governance theory, socio-technical ethics, and common-law principles of duty of care and foreseeability, the paper analyses how algorithmic decision architectures transform judgement into system output, explanation into documentation, and redress into procedural closure. Through a grounded case narrative situated in the United Kingdom public welfare context, it illustrates how appeals mechanisms and governance assurances can function to confirm system correctness while foreclosing substantive accountability after harm has occurred. The paper argues that explainability initiatives, while necessary, are insufficient to address this problem. Accountability requires answerability, discretion, and human authority to justify and revise decisions in light of their consequences. When these capacities are designed out of organisational systems, harm can occur without anyone being able to say, “I decided,” or “I will answer for this.” The paper concludes that accountability collapse is not an unintended by-product of automation, but an emergent property of governance choices, and that ethical responsibility in algorithmically mediated organisations must therefore be deliberately designed rather than assumed. This paper completes a three-part examination of ethical responsibility in algorithmically mediated organisations by focusing on accountability after harm.