CWI Cryptology Group

Abstract: The security of modern cryptosystems relies on computational assumptions, which may be challenged by the advent of large-scale quantum computing devices. While Shor's algorithm is known to break today's most popular public-key schemes, secret-key cryptosystems are generally expected to retain half of their pre-quantum bits of security. However, the precise advantage of quantum attacks cannot be determined without a dedicated analysis.
In this talk, we will focus on key-recovery attacks against block ciphers, including idealized constructions and actual designs. These attacks are often categorized in two scenarios, depending on the type of black-box access allowed to the adversary: either a classical query access, or a "quantum" query access where the black-box is part of the adversary's quantum algorithm. Attacks with classical queries, which are deemed more realistic, have so far complied with the rule of halving security levels. On the contrary, attacks with quantum queries can break some classically secure designs (Kuwakado & Morii, ISIT 2010), by exploiting a strong algebraic structure.
More recent results (Bonnetain et al., ASIACRYPT 2019) have shown that in some cases, an adversary making only classical queries can also be advantaged by the algebraic structure of the scheme. In particular, it will help to reduce its memory needs. But how far this advantage extends is an interesting open question.

Abstract: Cryptographic hash functions play significant roles in both theory and practice. Studying their security against quantum attacks is indispensable in the post-quantum era. This talk will overview recent progress on dedicated quantum collision attacks on concrete hash functions, including my results on AES-like hashes (Eurocrypt 2020) and SHA-2 (ePrint 2021/292), which are joint works with Yu Sasaki.
I will begin with basics such as the BHT algorithm and some important observations on dedicated collision attacks.