Alexander Viand
February 14, 2024 at 11:00 AM on Zoom / Soda Hall
Secure Computation: From Theory to Practice
Abstract: Secure Computation techniques such as Fully Homomorphic Encryption (FHE) are seeing increasing real-world deployment, protecting data in use by allowing computation over encrypted data. However, major obstacles to democratizing their use remain, since developing secure and efficient secure computation applications is complex and currently requires significant understanding of the underlying crypto. In order to enable wider adoption and unlock its full potential, we need to address this complexity and improve accessibility. At the same time, we need to consider how to safely deploy secure computation in the real world. Here, the same malleability that enables homomorphic computations also raises integrity issues, which have so far been mostly overlooked.In this talk, I will cover my work on making secure computation accessible to a broader audience via the development of tools and abstractions such as our end-to-end HE compiler framework, HECO, which allows non-expert developers to achieve performance levels usually reserved for state-of-the-art hand-tuned expert solutions. I will then discuss how we need to augment FHE with additional techniques such as Zero-Knowledge Proofs in order to deploy it in a broader range of settings, such as those with actively malicious adversaries. I will analyze existing approaches and identify capabilities, shortcomings and promising directions, providing experimental evaluations on realistic workloads. Finally, I will conclude with a discussion of current capabilities and potential future research directions.
Bio: Alexander Viand is a security & cryptography researcher at Intel Labs and a visiting scholar at UC Berkeley. In his research, he works with secure computation technologies including Fully Homomorphic Encryption, trying to make these techniques more accessible to non-experts by developing new systems, tools and abstractions. More generally, his interests include usable security and privacy, privacy enhancing technologies, and the interactions between these technologies and society. He received his PhD from ETH Zurich, where he was advised by Kenny Patterson (Applied Cryptography Group) and Anwar Hithnawi (Privacy Preserving Systems Lab).