Drake's Proposal: A Controlled Retreat, Not a Scramble
Drake's warning, posted to X on October 7, calls on the blockchain industry to "calmly begin planning for 'bunker mode.'" His core proposal is a gradual, voluntary migration: holders move remaining assets into addresses whose public keys have never been exposed onchain, starting with large holders and institutions rather than everyday users. Notably, Drake's immediate proposal doesn't require a new wallet format or a new cryptographic algorithm, addresses generated from the same seed phrase can serve the purpose, since the vulnerability lies specifically in exposed public keys, not in the underlying wallet technology itself.
The mechanics matter here. ECDSA secures standard Ethereum accounts and most Bitcoin wallets by using a private key to generate a signature, while the matching public key lets the network verify that signature without revealing the private key itself. Ethereum's own documentation confirms that an account which has only received funds, and never sent a transaction, has not exposed its public key onchain. The moment an account signs a transaction, though, its public key becomes recoverable from that signature, which is the exposure Drake is warning about. He repeatedly cautioned against panic throughout his post, warning that a poorly managed migration could introduce its own risks, user error during a rushed move is arguably a more immediate danger than the theoretical attack itself.
Buterin Agrees on the Risk, Not the Urgency
Vitalik Buterin's reaction adds useful nuance rather than amplifying alarm. He said he doesn't "recommend anyone scramble" to move funds today, while agreeing that AI-driven advances in mathematics deserve to be taken seriously, and that the industry should generally reduce its dependence on cryptographic assumptions that could prove weaker than currently believed. His concern isn't limited to elliptic-curve cryptography like ECDSA, Buterin specifically flagged that the concrete security of lattice-based cryptography could also face pressure from AI-assisted mathematical discovery over roughly the next two years, though he stopped short of claiming any lattice-based system has actually been broken. His practical guidance favors hash-based cryptographic designs where feasible and more conservative parameters for lattice-reliant systems, alongside his broader message: take the risk seriously, but don't let urgency create new mistakes.
Where OpenAI's Math Release Actually Fits In
Drake explicitly tied the timing of his warning to OpenAI's October 6 publication of mathematical research generated by an internal, unreleased frontier model, a public repository now containing 722 manuscripts across 372 result families, produced after the model was tested on roughly 4,000 research problems.
It's worth being precise about what that release actually was. OpenAI's announcement did not claim any attack against ECDSA, RSA, or cryptocurrency systems specifically, and the repository itself states that results are at varying stages of verification, not every manuscript includes formal proof verification, and some unverified results "could have issues." Drake's interpretation of what this pace of progress might eventually enable is considerably more aggressive than OpenAI's own framing; he wrote that "mathematical superintelligence is upon us" and questioned whether AI-assisted classical algorithms could find shortcuts against elliptic-curve cryptography before quantum computers become capable of the attack at meaningful scale. That remains his personal risk assessment, not a demonstrated cryptographic break, a distinction several outlets covering this story have been careful to preserve.
Ethereum's Post-Quantum Groundwork Was Already Underway
This warning doesn't arrive in a vacuum. Ethereum has been preparing for cryptographic transition for some time: its Post-Quantum Security team, formed in January 2026, has been testing hash-based validator signatures, new proof systems, and post-quantum interoperability across multiple client teams. The current roadmap centers on leanXMSS, a hash-based signature design meant to eventually replace quantum-vulnerable validator signatures, paired with leanVM for efficiently aggregating larger post-quantum signatures, targeting core infrastructure readiness around 2029, though Ethereum describes that date as a planning target rather than a firm commitment. Separately, EIP-8141 would eventually give individual accounts more flexibility in how transactions are authenticated, potentially allowing a quantum-resistant signature method without requiring a full move to a new address format. Outside Ethereum, NIST finalized a hash-based post-quantum signature standard (SLH-DSA) back in 2024, and institutional custody providers including BitGo and Coinbase have separately begun testing post-quantum approaches of their own.
