Black Alps 2024

Abstract

In 2021 Li and Miccianci showed that the notion of IND-CPA security is not always sufficient to provide passive security in the context of approximate homomorphic encryption. As a solution, they introduced the notion of IND-CPA-D security but left open the question on how to achieve it efficiently in practice. Shortly after, it was shown that all mainstream lattice based FHE schemes actually end up being vulnerable to the same attack and that they also need to satisfy the notion of IND-CPA-D security in these scenarios. Since then the FHE community has been divided and caught in an iterative process on how to achieve IND-CPA-D security in a practical and provable way, with most of the proposed solutions ending up being either broken or impractical. Can IND-CPA-D security be achieve in a way that is both practical and provably secure?

Jean-Philippe Bossuat

Cryptography Research Scientist & Software Specialist

My research interests include applied lattice based cryptography, privacy preserving machine learning and secure analytics.