Charles Hoskinson Slams Vitalik Buterin's AI Cryptography Warning as Panic

Published October 9, 2026, 10:25 UTC. Vitalik Buterin posted his warning on Oct. 7, 2026; Charles Hoskinson responded with his rebuttal on Oct. 8, 2026.

The loudest argument in crypto this week is not about prices. It is about whether artificial intelligence could break the cryptography that secures blockchains, and it has put two of the industry's most prominent founders in open disagreement. On October 7, Ethereum co-founder Vitalik Buterin warned on X that AI-accelerated mathematics could erode the security of lattice-based cryptography within about two years. On October 8, Cardano founder Charles Hoskinson fired back in a long X post, calling Buterin's statements panic and baseless fears about the backbone of the post-quantum security paradigm.

Buterin's warning

Buterin's post extended a debate that began days earlier, when Ethereum researcher Justin Drake warned the industry to prepare for a "bunker mode" scenario in which AI could accelerate mathematical attacks on the elliptic-curve cryptography (ECDSA) used to authorize transactions. The conversation sharpened after OpenAI published new machine-generated mathematical proofs, formalized in the Lean proof assistant, produced by an internal frontier model on October 6. OpenAI reported no attack on blockchain cryptography, but the event underscored how quickly AI can advance mathematical discovery.

Buterin argued that lattice-based systems, the leading candidates for post-quantum signatures, deserve less confidence than the industry gives them. Lattices underpin ML-DSA, the digital signature standard the U.S. National Institute of Standards and Technology standardized in August 2024 as believed secure even against large-scale quantum computers. Buterin's point was about unknown attacks, not known ones. He drew on the history of integer factorization, where the general number field sieve made attacking RSA far easier than naive approaches suggested. "If AI will bring us 50 years of math in 2 years," he wrote, the advances could include breakthroughs that substantially improve attacks on lattice cryptography. He said Ethereum's lean roadmap has moved toward hash-only signatures for that reason, advised preferring hash-based tools, and suggested multiplying lattice key sizes by 10. Crucially, he told holders not to rush to move funds, writing that he did not recommend anyone "scramble to move their funds to new wallets today."

Hoskinson's rebuttal

Hoskinson rejected the argument point by point. He said lattice cryptography has been subjected to deep research and rigorous testing for roughly 40 years, with researchers constantly refining security assumptions as attack methods evolved. He called Buterin's position highly speculative and criticized it for lacking mathematical substantiation. He also accused Buterin of "bag-holding too much hash-based crypto research," arguing that hashes are not immune either: MD5 and SHA-1 were both broken, and Ethereum's own Poseidon hash function is so mathematically structured that the Ethereum Foundation funded a bounty to attack it.

The harm argument

Hoskinson went further than a technical rebuttal. He warned that Buterin's stance could do grave harm to the cybersecurity industry as a whole. Discouraging people from using lattice cryptography, he said, could undermine the ongoing effort to secure internet communications against future quantum computers and hamper the overall transition from classical cryptography, endangering sensitive information that could be compromised later. In his view, seeding doubt about NIST-standardized lattice signatures risks more damage than the hypothetical AI breakthrough Buterin described.

The clash matters because it exposes a real divide in how crypto plans for cryptographic risk. Buterin wants the industry to hedge against not just quantum-vulnerable cryptography but "AI-vulnerable" cryptography, treating mathematical structures as suspects until proven otherwise. Hoskinson wants the industry to trust decades of adversarial testing and to avoid a panic-driven migration that could itself create operational risks. Neither side pointed to a working exploit, and no practical attack against ECDSA or ML-DSA has been publicly demonstrated.

This article is news reporting and is not investment advice.

Sources

• Vitalik Buterin, X post (Oct. 7, 2026), primary source
• Charles Hoskinson, X post (Oct. 8, 2026), primary source
• U.Today, Cardano Founder Slams Vitalik Buterin Over Crypto Security (Oct. 8, 2026), secondary source
• CCN, Vitalik Buterin Warns AI Could Weaken Cryptography in 2 Years (Oct. 8, 2026), secondary source
• Coinpedia, Cardano Founder Accuses Ethereum Co-Founder of 'Bag-Holding' Over AI Cryptography Warning (Oct. 8, 2026), secondary source