Is ML-Based Cryptanalysis Inherently Limited? Simulating Cryptographic Adversaries via Gradient-Based Methods
Avital Shafran
Abstract:
Given the recent progress in machine learning (ML), the cryptography community has started exploring the applicability of ML methods to the design of new cryptanalytic approaches. While current empirical results show promise, the extent to which such methods may outperform classical cryptanalytic approaches is still somewhat unclear. In this work, we initiate exploration of the theory of ML-based cryptanalytic techniques, in particular providing new results towards understanding whether they are fundamentally limited compared to traditional approaches. We introduce a unifying framework for capturing both ``sample-based'' and ``gradient-based'' adversaries. Within our framework, we establish a general feasibility result showing that any sample-based adversary can be simulated by a seemingly-weaker gradient-based one.
Bio:
Avital is a PhD student at the Hebrew University, advised by Prof. Shmuel Peleg and Prof. Gil Segev. She is interested in the security of machine learning and the intersection between machine learning and cryptography.