Neurohash Assessment of Quantum Attack Resistance

This work is a direct continuation of the study [1] and is devoted to assessing the resistance of the neural network hash function NeuroHash to quantum attacks. Unlike the first publication, which focused on classical cryptography and NIST certification, this paper presents for the first time an analysis using Grover's algorithm. A simulation of quantum preimage search was performed on a classical computer using a developed C++ test bench. A theoretical assessment of the complexity of breaking NeuroHash in the post-quantum era is carried out. It is shown that even taking into account quadratic acceleration, the attack complexity is 2²⁵⁶ operations, which guarantees quantum resistance. Graphs of the dependence of success probability on the number of iterations and a comparative analysis with existing standards are presented.