According to CoinMeta, a16z crypto and Stanford cryptography professor Dan Boneh among others pointed out that the latest mathematical findings of OpenAI have not proven that existing cryptographic systems such as elliptic curves are more vulnerable, nor have any new algorithms been discovered that can efficiently crack core cryptographic problems. Compared to directly cracking cryptography, the more practical impact of AI is to significantly accelerate the verification of cryptographic assumptions, the discovery of vulnerabilities, and software security testing, greatly shortening what used to be review cycles that could take several years. At the same time, post-quantum migration and cryptographic agility remain key to addressing future uncertainties associated with AI and quantum computing.