A warning from an Ethereum Foundation researcher has split opinion across the crypto industry this week. Justin Drake says artificial intelligence could find a way to break the mathematics protecting bitcoin and ether wallets. In the worst case, he says, that could happen in months rather than years. He wants the industry to prepare for what he calls “bunker mode.”
Drake’s concern centers on elliptic-curve cryptography, the signature system that proves ownership of coins on Bitcoin and Ethereum. Much of the debate over future threats has focused on quantum computers. Drake argues AI-driven mathematical discovery could arrive first.
The timing follows OpenAI’s release on Tuesday of 722 mathematical manuscripts produced by an unreleased model tested on roughly 4,000 research problems. Security developments over the summer also fed the alarm. In late July, a volunteer group called the Bitcoin Red Team used AI models to scan 390 Bitcoin software projects in about 27 hours. It logged nearly 5,000 possible flaws, 85 of them rated critical. Core Lightning’s developers later issued an emergency warning after AI-generated bug reports uncovered real vulnerabilities in their code.
Drake’s advice is aimed at sophisticated holders. He suggests moving funds to addresses whose public keys have never appeared on-chain, which denies an attacker the starting point the attack would need. Exposure is already widespread. Glassnode data highlighted by CoinDesk shows more than 6 million bitcoin sit behind exposed public keys.
Vitalik Buterin took a more measured line. He said the risks from AI-accelerated math deserve serious attention and backed reducing public-key exposure where practical. But he told holders not to scramble, arguing that mistakes during a wallet migration are a more immediate danger than any demonstrated break. He also flagged that lattice-based cryptography, including the ML-DSA signature scheme, could see its security margins erode over the next two years. That matters because lattice schemes are among the leading post-quantum candidates.
Others were dismissive. Samson Mow, CEO of Jan3, told followers there was no reason to panic over what he called silly claims from an Ethereum researcher. Commentators also noted that institutions face a harder task than individuals. Large custodians have to migrate vast numbers of addresses under compliance requirements.
The warning landed on a weak day for the market. Bitcoin dipped to about $82,300 before recovering to around $82,800, roughly 4% below Tuesday’s high near $86,600. Analysts attribute the slide mainly to rate-hike expectations, rising oil prices and fund outflows, so the AI debate was likely a minor factor.
No break of the cryptography has been demonstrated. Developers are expected to keep pushing for wider adoption of quantum- and AI-resistant signature schemes. For now, the practical steps are narrow: avoid reusing addresses and follow protocol-level migration guidance as it emerges.
