Multi-Party Computation as a Privacy-Enhancing Technology

Abstract

Data sharing through data marketplaces, which rely on a Trusted Third Party (TTP), can benefit businesses and society. However, many companies and consumers are increasingly reluctant to share data due to mounting concerns over data control and privacy. Emerging privacy-enhancing technologies (PETs) like Multi-Party Computation (MPC), which enables joint computation to generate insights while keeping the input data private, could address data sharing barriers due to its differences with the traditional data sharing approach relying on a TTP. Thus, MPC could challenge the current understanding of why and how businesses and consumers share data. Nevertheless, whether businesses and consumers would be more willing to share data with MPC in place is unclear, as less attention is given to the socio-technical implications of MPC on data sharing decisions in data marketplaces and its antecedents. This research aimed to theorize the socio-technical implications of MPC on sharing through data marketplaces, by investigating how MPC potentially impacts data sharing antecedents by businesses and individuals. We do so through a mixed-method research design focusing on the automotive industry. Based on interviews with 15 MPC experts, which were structured using a Unified Business Model framework, we explored value propositions enabled by MPC use in data marketplaces. These value propositions allow MPC to potentially impact control, privacy, trust, and risks as antecedents of data sharing decisions in data marketplaces. Subsequently, we interviewed 23 automotive industry experts to investigate the potential impact of MPC use in data marketplaces on control, trust, and risks as antecedents of business data sharing. We then conducted an experiment via an online crowdsourcing platform with 1457 participants to investigate the potential impact of MPC use in data marketplaces on control, privacy, trust, and risks as antecedents of consumer data sharing. In this way, we contribute to the socio-technical understanding of MPC beyond technical perspectives. At the same time, we also demonstrate the relevance of MPC to practitioners by pointing out key aspects that should be considered while exploring the possibility of implementing MPC. Furthermore, this research provides a foundation for future studies on understanding the socio-technical implications of MPC on data sharing decisions.