Title:

Privacy-Preserving Federated Analytics with Multiparty Homomorphic Encryption

 

Abstract:

Collecting, analyzing and sharing data has become of crucial importance in numerous domains. However, in many cases, the data are sensitive and personal, and sharing them is often difficult due to legitimate privacy concerns and necessary data-sharing regulations. In domains where the data are extremely sensitive, e.g., biomedical research, the collected data are often not shared across different institutions and remain in silos. This hinders research by restricting access to large and diverse datasets, which is required for discovering new scientific and clinical insights.
In this talk, I will introduce a new framework for privacy-preserving federated analysis based on multiparty homomorphic encryption. I will explain how our framework overcomes privacy concerns, notably in the context of biomedical research, by leveraging state-of-the-art cryptographic techniques to ensure end-to-end data confidentiality and to enable fine-grained access control to the analysis results. Contrary to alternative approaches, our solution does not introduce noise in the computation for privacy protection, and it enables a large number of entities to collaborate while locally keeping their private data. I will demonstrate our framework’s applicability by replicating, in a federated and privacy-preserving manner, a range of essential biomedical analysis tasks, including genome-wide association studies. Our framework has the potential to accelerate research by enabling privacy-preserving access to siloed data and unlocking new collaborative studies.

 

Speaker:

David Froelicher

Post-Doctoral Researcher
MIT

and

The Broad Institute of MIT and Harvard

 

Bio:

David Froelicher is a Post-Doctoral Researcher working with Prof. B. Berger in the Computer Science and Artificial Intelligence Laboratory (CSAIL) at the Massachusetts Institute of Technology (MIT) and with Dr. H. Cho at the Broad Institute of MIT and Harvard. David focuses on privacy-preserving federated analytics and genomic privacy. He designs new secure and distributed solutions by building on top of applied cryptography, notably using homomorphic encryption and secure multiparty computation. David received his PhD from the Ecole Polytechnique Fédérale de Lausanne (EPFL) for his work with Prof. Jean-Pierre Hubaux at the Laboratory for Data Security (LDS) and Bryan Ford at the Decentralized and Distributed Systems Laboratory (DeDiS).