Adopting and Scaling Secure Systems with Distributed Trust

Vivian Fang

EECS Department
University of California, Berkeley
Technical Report No. UCB/EECS-2025-106
May 16, 2025

http://www2.eecs.berkeley.edu/Pubs/TechRpts/2025/EECS-2025-106.pdf

Distributed trust is an emerging design pattern that enables building secure systems with strong privacy and integrity guarantees. Yet, adopting these systems in practice poses significant hurdles, from requiring specialized cryptographic expertise to scaling performance and materializing independent trust among parties. To this end, we introduce a suite of works aimed at making secure systems with distributed trust more practical, scalable, and ultimately deployable. On usability, we develop CostCO, an automatic cost-modeling framework for secure multi-party computation protocols, and LegoLog, a configurable transparency log system that automatically generates logs tailored to specific application workloads. We explore scaling oblivious storage in Snoopy, a system capable of scaling throughput linearly with the number of machines without sacrificing security. Finally, we present SVR3, a practical solution for secret key recovery that distributes trust across heterogeneous hardware enclaves in order to protect secrets at scale for hundreds of millions of users.

Advisor: Raluca Ada Popa

\"Edit"; ?>

BibTeX citation:

@phdthesis{Fang:EECS-2025-106,
    Author = {Fang, Vivian},
    Title = {Adopting and Scaling Secure Systems with Distributed Trust},
    School = {EECS Department, University of California, Berkeley},
    Year = {2025},
    Month = {May},
    URL = {http://www2.eecs.berkeley.edu/Pubs/TechRpts/2025/EECS-2025-106.html},
    Number = {UCB/EECS-2025-106},
    Abstract = {Distributed trust is an emerging design pattern that enables building secure systems with strong privacy and integrity guarantees. Yet, adopting these systems in practice poses significant hurdles, from requiring specialized cryptographic expertise to scaling performance and materializing independent trust among parties. To this end, we introduce a suite of works aimed at making secure systems with distributed trust more practical, scalable, and ultimately deployable. On usability, we develop CostCO, an automatic cost-modeling framework for secure multi-party computation protocols, and LegoLog, a configurable transparency log system that automatically generates logs tailored to specific application workloads. We explore scaling oblivious storage in Snoopy, a system capable of scaling throughput linearly with the number of machines without sacrificing security. Finally, we present SVR3, a practical solution for secret key recovery that distributes trust across heterogeneous hardware enclaves in order to protect secrets at scale for hundreds of millions of users.}
}

EndNote citation:

%0 Thesis
%A Fang, Vivian
%T Adopting and Scaling Secure Systems with Distributed Trust
%I EECS Department, University of California, Berkeley
%D 2025
%8 May 16
%@ UCB/EECS-2025-106
%U http://www2.eecs.berkeley.edu/Pubs/TechRpts/2025/EECS-2025-106.html
%F Fang:EECS-2025-106