It is not (only) about privacy: How multi-party computation redefines control, trust, and risk in data sharing

Firms are often reluctant to share data because of mistrust, concerns over control, and other risks. Multi-party computation (MPC) is a new technique to compute meaningful insights without having to transfer data. This paper investigates if MPC affects known antecedents for data sharing decisions: control, trust, and risks. Through 23 qualitative interviews in the automotive industry, we find that MPC (1) enables new ways of technology-based control, (2) reduces the need for inter-organizational trust, and (3) prevents losing competitive advantage due to data leakage. However, MPC also creates the need to trust technology and introduces new risks of data misuse. These impacts arise if firms perceive benefits from sharing data, have high organizational readiness, and perceive data as non-sensitive. Our findings show that known antecedents of data sharing should be specified differently with MPC in place. Furthermore, we suggest reframing MPC as a data collaboration technology beyond enhancing privacy.