Staying at the Edge of Privacy: Edge Computing and Impersonal Extraction

From self-driving cars to smart city sensors, billions of devices will be connected to networks in the next few years. These devices will collect vast amounts of data which needs to be processed in real-time, overwhelming centralized cloud architectures. To address this need, the industry seeks to process data closer to the source, driving a major shift from the cloud to the ‘edge.’ This article critically investigates the privacy implications of edge computing. It outlines the abilities introduced by the edge by drawing on two recently published scenarios, an automated license plate reader and an ethnic facial detection model. Based on these affordances, three key questions arise: what kind of data will be collected, how will this data be processed at the edge, and how will this data be ‘completed’ in the cloud? As a site of intermediation between user and cloud, the edge allows data to be extracted from individuals, acted on in real-time, and then abstracted or sterilized, removing identifying information before being stored in conventional data centers. The article thus argues that edge affordances establish a fundamental new ‘privacy condition’ while sidestepping the safeguards associated with the ‘privacy proper’ of personal data use. Responding effectively to these challenges will mean rethinking person-based approaches to privacy at both regulatory and citizen-led levels.