Charles Hoskinson rebutted Vitalik Buterin's concerns about lattice-based cryptography's security and the post-quantum migration path.
Article
【Cardano Founder Rebuts Vitalik: Lattice-Based Cryptography Should Not Be Abandoned Due to AI Concerns】October 9, Cardano founder Charles Hoskinson posted a rebuttal to Vitalik Buterin's 2 days ago regarding post-quantum cryptography. Vitalik believes that AI accelerated mathematical research could weaken the concrete security of lattice-based cryptography (including ML-DSA、FHE) and said that Ethereum's Lean roadmap has therefore shifted to hash-only signatures; he also suggested increasing lattice-scheme key lengths by about 10 times and keeping encrypted notes in privacy protocols off-chain whenever possible.
Hoskinson said this assessment lacks a citable attack and cost model. The number field sieve cannot simply be analogized to lattice problems, and the parameters of ML-KEM、ML-DSA already account for decades of sieve and enumeration attacks; quantum attacks on ideal lattices have also not reached current module-lattice schemes. By contrast, hashes such as Poseidon, which Ethereum uses, have stronger algebraic structure themselves. Hoskinson believes that abandoning the deployed hybrid ML-KEM before an attack appears would instead keep traffic on classical cryptography for longer, intensifying the “harvest now, decrypt later” risk.
Key points
01
October 9, Charles Hoskinson posted a rebuttal to Vitalik Buterin 2 days ago regarding post-quantum cryptography.
02
Vitalik Buterin believes that AI accelerated mathematical research could weaken the concrete security of lattice-based cryptography, including ML-DSA、FHE.
03
The original text states that Vitalik Buterin said Ethereum's Lean roadmap has therefore shifted to hash-only signatures.
04
Vitalik Buterin suggested increasing lattice-scheme key lengths by about 10 times and keeping encrypted notes in privacy protocols off-chain whenever possible.
05
Charles Hoskinson said the above assessment lacks a citable attack and cost model.
06
Charles Hoskinson said that ML-KEM and ML-DSA's parameters already account for decades of sieve and enumeration attacks.
07
Charles Hoskinson believes that abandoning the deployed hybrid ML-KEM before an attack appears would keep traffic on classical cryptography for longer, intensifying the “harvest now, decrypt later” risk.
AI-assisted interpretation
The following is analysis, separate from reported facts. Verify important claims independently.
The two sides are debating how to address the cryptographic risks that future quantum computing and AI could bring. Vitalik favors using hash-only signatures and increasing the security margin of lattice schemes; Hoskinson believes that no specific attack method or cost has currently been provided, and that deployed hybrid ML-KEM schemes should not therefore be abandoned. The original text presents a disagreement over technical positions; it does not confirm that any cryptographic scheme has been compromised.
Why it matters to readers
This disagreement concerns how blockchain systems should choose post-quantum security solutions in the future and whether technical paths should be adjusted before specific attacks emerge. Different choices could affect key lengths, on-chain data arrangements, and the method of migrating from classical cryptography.
Beginners should understand this as two industry figures making different judgments about preventing future cryptographic risks—not as a security vulnerability having occurred in Cardano or Ethereum, nor as a confirmed protocol change.
Risks and unknowns
The source material does not include the original posts by Charles Hoskinson or Vitalik Buterin, so the specific wording and context cannot be verified.
The source material does not provide a reproducible attack, cost model, technical paper, or independent security assessment.
It cannot be confirmed whether Ethereum's Lean roadmap has formally decided to shift to hash-only signatures.
Related Developments
Loading event timeline…
The specific deployment scope and status of the hybrid ML-KEM referred to in the text cannot be confirmed.
Existing materials do not indicate that Cardano or Ethereum has experienced the relevant security vulnerability.
Related concepts
lattice-based cryptography
This term is not in the glossary yet. Browse related concepts in the glossary.