Decision-making by proxy: A contextual study of data intermediaries

Ensuring privacy in digital ecosystems presents significant challenges. Individual privacy-related decision-making has been widely criticised as insufficient. The mobility context, where computing is typically ubiquitous, makes this type of decision-making also operationally infeasible. This study examines whether innovative data governance strategies offer a promising approach. Drawing on contextual integrity theory, this paper examines the perceived appropriateness of delegating privacy-related decision-making to different types of data intermediaries (data cooperatives, AI agents, and experts) in mobility contexts. Specifically, a stratified, full-factorial vignette study ( N  = 600) is conducted. The vignettes systematically varied across four contextual factors: sender, information type, purpose, and duration of data access approval. This design allows for the examination of how contextual factors influence preferences for data intermediary-based decision-making in mobility settings. Findings show that scenarios in which data cooperatives acted as the decision-makers were consistently rated as more appropriate than those involving AI agents or experts, particularly in high-stakes contexts such as traffic management and ambulance coordination. Nevertheless, scenarios involving individual control received the highest overall appropriateness ratings. Respondents’ preferences were context-dependent: they showed greater acceptance when the purpose of sharing aligned with public benefit, and when the access duration was either short (e.g., one day) or revocable (e.g., ‘until revoked’). The results underscore the relevance of hybrid governance models that combine individual control with the ability to delegate decision-making to trusted intermediaries, particularly in mobility contexts. Together, these contributions could aid in developing context-sensitive privacy frameworks that align with societal values in ubiquitous computing environments.