Privacy-Preserving Large-Scale Computation: Foundations and Applications

Objetivo

In In the era of exponential growth in available data, privacy and security have become paramount concerns for a wide range of domains, including healthcare, financial services, and many more. Classical tools developed in the cryptography community, e.g. the ability to perform computation on encrypted data without revealing the underlying information, could have the potential to address these challenges. However, most results do not fit modern large-scale distributed settings due to communication and storage limitations. A recent line of work attempts to bridge this gap, yet, existing solutions are still highly unsatisfactory since they assume either strong trusted setup assumptions, rely on strong and non-standard computational assumptions, or they are extremely theoretical.

This research proposal aims to overcome the above drawbacks and systematically identify and design secure computation tools that can be effectively used within modern large-scale applications. To this end, we will address three main challenges: (1) understand and design distributed large-scale building blocks (such as broadcast protocols), (2) equip large-scale algorithms with privacy guarantees, and (3) increase the usability and practical usefulness of privacy-preserving tools in large-scale settings. A successful execution of the proposed research will capitalize on and develop new ideas in three key areas: cryptography, distributed computing, and algorithms.

This research holds immense potential for applications in various domains. For instance, it could enable secure analysis of patient data across healthcare providers, improving diagnosis and treatment planning while preserving privacy. Similarly, in financial services, it could revolutionize fraud detection by collaborative analysis without compromising customer confidentiality. Ultimately, this research's goal is to design novel methods for privacy-preserving data processing, allowing new applications benefiting society as a whole.