On October 8, Ethereum co-founder Vitalik Buterin warned that AI-accelerated mathematics could break lattice-based cryptography, the foundation of post-quantum security standards. Hours later, Cardano founder Charles Hoskinson rejected the warning, characterizing it as panic and suggesting Buterin's position reflected financial interests in hash-based alternatives.
Buterin argued that AI could discover weaknesses in lattice cryptography that researchers have overlooked for decades. He cited RSA encryption as precedent, noting that defenses once considered secure eventually fell to faster attacks. He recommended Ethereum move toward hash-only signatures and suggested increasing lattice key sizes by 10 times as a precaution.
Hoskinson's Counter-Arguments
Hoskinson challenged Buterin's reasoning across multiple points. He noted that lattice-based cryptography has withstood 40 years of attacks from leading researchers and that current standards are already sized to resist known threats. He also pointed out that hash functions are not immune to breaks, citing MD5 and SHA-1 as examples. Regarding the proposed 10-fold key expansion, Hoskinson called it "numerology," arguing that a 20 percent stronger attack would require approximately 25 percent larger keys.
Hoskinson raised concerns about the broader implications of Buterin's position. Lattice-based ML-KEM is already deployed in browsers and messaging applications to protect against future decryption of currently captured encrypted data. Casting doubt on lattice security, Hoskinson argued, could delay adoption of stronger defenses in mainstream internet traffic.
He also questioned Buterin's track record, noting that past recommendations had steered engineers away from concepts like Plasma and Casper, only for the industry to revisit them later.
Neither founder cited evidence of an actual AI-discovered attack on lattice or hash-based cryptography. The debate centers on whether the cryptocurrency sector should preemptively guard against a potential breakthrough or continue relying on defenses that have proven effective over decades.
