Abstract This article contributes a coherent framework to the rich literature emerging in the field of algorithmic governance while also resolving conflicting understandings. Tracing the history of algorithmic governance to the broad architecture of the universal Turing Machine, the article identifies a common thread of critical concern in the literature on algorithmic governance: the growing institutional capabilities to move contestable issues to a space of reduced negotiability, raising questions of social asymmetry, inequity, and inequality. Within the social context of algorithmic governance, the article highlights three general areas of concern where the social negotiability of processes is threatened: the problem of power (surveillance), discrimination (social bias), and identification (system identity).