Charles Hoskinson, founder of Cardano, challenged Vitalik Buterin's recent warnings about artificial intelligence accelerating breakthroughs in cryptographic attacks. In a public statement on October 9, Hoskinson argued that Buterin's skepticism toward lattice-based cryptography undermines decades of established security research. He emphasized that lattice-based systems, including those standardized by NIST in 2024, have undergone rigorous analysis addressing known vulnerabilities. The critique follows Buterin's suggestion that AI-assisted mathematical discoveries could expose weaknesses in post-quantum cryptographic designs, particularly those relying on lattice structures.

Hoskinson rejected the notion that AI could rapidly advance cryptanalysis capabilities against lattice problems, noting that no comparable mechanism to the general number field sieve exists for these systems. He referenced over 40 years of research in lattice reduction and sieving algorithms, which have incrementally improved attack methods without achieving a decisive breakthrough. The Cardano founder also dismissed Buterin's proposal to increase key sizes by tenfold as 'numerology,' asserting that security parameters must reflect measurable improvements in attack efficiency rather than speculative scenarios.

The dispute centers on post-quantum cryptographic strategies. While Ethereum's roadmap incorporates hash-based signatures and hash functions like Poseidon for zero-knowledge applications, Hoskinson questioned their resilience against unforeseen mathematical advances. He pointed to historical vulnerabilities in MD5 and SHA-1 as evidence that hash functions can contain structural weaknesses, though he acknowledged these failures do not invalidate modern constructions like SHA-256. The Ethereum Foundation's funding of Poseidon research, including analyses using Gröbner bases, was cited as an example of ongoing scrutiny applicable to multiple cryptographic approaches.

Hoskinson warned that abandoning lattice research in favor of alternative methods could limit cryptographic options for developers building privacy-preserving systems and secure communications infrastructure. He maintained that NIST-standardized lattice-based algorithms, such as ML-KEM for encryption and ML-DSA for signatures, already incorporate security margins accounting for known attack vectors. The critique underscores broader tensions in the cryptocurrency industry regarding the path toward quantum-resistant cryptography, with implications for blockchain security and cross-chain interoperability.