Secure Key Management for Multi-Party Computation in MOZAIK

Abstract

The immense growth of data from the proliferation of Internet of Things (IoT) devices presents opportunities and challenges for privacy engineering. On the one hand, this data can be harnessed for personalized services, cost savings, and environmental benefits. On the other hand, (new) legislation must be complied with and privacy risks arise from collecting and processing of such data. Distributed privacy-preserving analytics offers a promising solution, providing insights while also protecting privacy. However, this approach has new challenges and risks, such as key management and confidentiality. When designing a data marketplace which offers distributed privacy-preserving analytics, the key management comes with different threats, which require a solution adapted to the distributed architecture.

In this context, the paper presents a comprehensive, end-to-end secure system called MOZAIK for privacy-preserving data collection, analysis, and sharing. The article focuses on the key management aspect of the secure multi-party computation (MPC) component in a distributed privacy-preserving analytics architecture and the specific challenges created by introducing MPC. The proposed solution involves temporary storage of (symmetric) key shares and public-key encryption schemes to ensure secure key management for privacy-preserving computation. Our solution has the potential to be applied in other MPC-based setups, making it a valuable addition to the field of privacy engineering. By addressing key management challenges and risks, MOZAIK enhances data protection while enabling valuable insights from IoT data.