Vitalik Buterin: take AI’s threat to crypto cryptography seriously — but do not scramble

Ethereum co-founder Vitalik Buterin weighed in on October 7, 2026, after a warning that artificial intelligence could undermine the signatures that secure Bitcoin and Ethereum wallets before quantum computers do. His line was caution without panic: prepare, do not rush a mass wallet migration.

The exchange landed as Bitcoin traded below $83,000 and the warning split the industry. CoinDesk reported the slide and the debate the same morning.

What Buterin actually said

Buterin was answering Ethereum Foundation researcher Justin Drake, who on Wednesday called on the industry to plan for what he named “bunker mode.” Drake’s recommendation was a controlled move of assets to fresh addresses whose public keys stay hidden behind a hash. He argued it is now reasonable to prepare for the elliptic curve digital signature algorithm (ECDSA) — the scheme behind most crypto wallets — being broken before “Q-Day,” the point at which quantum computers can defeat widely used public-key cryptography. In a worst case, he put that breakthrough in months rather than years, meaning a private key recovered in about a week on a large GPU cluster.

Buterin did not endorse a rush. On X he wrote: “I don’t recommend anyone scramble to move their funds to new wallets today. But we should take the risks to cryptography from AI-accelerated math seriously, and minimize our exposure to not just quantum-vulnerable cryptography, but also potentially AI-vulnerable cryptography.”

Cointelegraph and The Block both led with that distinction: the risk is real enough to plan for; a botched migration is a nearer threat than a theoretical break.

Why the warning landed now

Drake tied the call to OpenAI’s release on Tuesday of 722 mathematical results across algebra, theoretical computer science, and logic. Cointelegraph also noted OpenAI’s earlier claim that a team of 10,000 autonomous agents solved the Navier-Stokes problem in 88 hours. Drake’s point was structural: “Curves carry rich structure, with room for fancy tricks like Schoof, Frobenius, pairings. (By contrast, hashes are designed to minimize algebraic structure.)”

News.Bitcoin.com carried the same debate in French, including Drake’s call to “calmly begin planning” for bunker mode and Buterin’s refusal to tell holders to move funds today.

Lattices, hashes, and Ethereum’s lean roadmap

Buterin widened the frame past ECDSA. The working assumption, he wrote, has been “elliptic curves broken, hashes safe, lattices safe.” He now thinks lattice schemes — including ML-DSA and fully homomorphic encryption — “will take serious hits from the next two years of AI math.” His analogy is factoring: naive methods scale like 2^{n/2}, but the number-field sieve cut that to roughly 2^{O(n^{1/3})}, which is why RSA keys are hundreds of bytes, not 64. If AI finds a similar hidden structure in lattices, he argued, keys may need to be about ten times larger to hold the same safety margin.

That is a main reason, he said, Ethereum’s lean roadmap has spent the past year going “hash-only”: no lattices, no ML-DSA, no Falcon, and hash-based signatures such as WOTS or SPHINCS. Public-key encryption cannot be built from hashes alone, he added, so the harder problem sits outside blockchains — messaging, anonymizing networks, and web access.

His practical list, in his own words:

• Prefer hash-based constructions where they are possible.
• Be “much more paranoid” on lattice parameter sizes.
• For privacy protocols, “strongly favor NOT putting encrypted notes onchain.”
• If it is easy, keep funds in addresses that have never signed a transaction.
• For multisig, confirm off-chain so signer public keys are not exposed.

He also wrote the line The Block highlighted: botched migrations have cost him more than hacks. “If it’s not difficult for you, keeping your funds in addresses which have not yet been used to make a transaction is a good idea. If it’s easy for you, do it. But be careful about migrations; I personally have lost more money in botched migrations than I have lost in all hacks combined.”

The split

Not everyone accepted the timeline. The Block reported that Yehuda Lindell, head of cryptography at Coinbase, said there is “no evidence whatsoever” that the hardness assumptions behind elliptic-curve cryptography have been weakened. A Solana-aligned commenter argued Solana users do not need bunker mode. CoinDesk quoted Jan3 CEO Samson Mow telling followers there was no need to panic “because an Ethereum researcher is saying silly things.”

Dragonfly’s Haseeb Qureshi took the other side, calling Drake’s note “a very sober call” and framing the risk as conventional mathematics overturning unproven hardness assumptions, not quantum hardware.

The disagreement is about urgency, not about the existence of a research problem. Drake and Buterin are arguing from the pace of AI math and from the algebraic structure of curves and lattices. Critics are arguing that no concrete break of ECDSA has been shown, and that a rushed user migration is itself an operational failure mode.

A recent post from his official account

Buterin’s verified account is @VitalikButerin. The October 7 post (more than 1.5 million views) is the source of the quotes above. The opening, in full:

“I don’t recommend anyone scramble to move their funds to new wallets today. But we should take the risks to cryptography from AI-accelerated math seriously, and minimize our exposure to not just quantum-vulnerable cryptography, but also potentially AI-vulnerable cryptography.”

Three weeks earlier, on September 17, Decrypt reported a different Buterin line on the same theme: “It’s an increasingly common take that AI hacking means cybersecurity is doomed. I disagree. I think cybersecurity is naturally defense-favoring once people get their shit together.” He argued AI could also prove that software is secure. The October post does not retract that. It narrows the claim: defense can win, but lattice parameters and exposed public keys should not be treated as settled.

Who is Vitalik Buterin?

Vitalik Buterin is the co-founder of Ethereum and its most visible protocol writer. He proposed the network in late 2013 and remains the reference voice on its research roadmap, including the lean, hash-based designs he cited this week. He posts technical arguments directly on X under @VitalikButerin rather than through a corporate communications team. In the September Decrypt piece he said roughly 90 percent of his net worth is in crypto, which is why the migration warning is partly autobiographical: he has lost more to bad key moves than to hacks.

What readers should watch

Nothing in these statements is a confirmed break of ECDSA or of lattice cryptography. The claims are about preparation under a faster math timeline. The facts that are on the record are the posts, the OpenAI math release Drake cited, Ethereum’s stated hash-only direction, and the public disagreement from Coinbase’s cryptography lead and others.

Related reading on Cryptopress

• Cryptocurrency: The Comprehensive Technical Guide (2026 Edition) — how wallet cryptography, hashes, and institutional Bitcoin fit the 2026 stack.