KOSUNI CRYPTO
Korea's Crypto Pulse, in English
$BTC $ETH $XRP $SOL $DOGE
News
Neutral

Justin Drake Warns Crypto Industry to Prepare for 'Bunker Mode' Amid AI Math Breakthrough

Published October 8, 2026 6:26 PM · 6 views $BTC
Justin Drake Warns Crypto Industry to Prepare for 'Bunker Mode' Amid AI Math Breakthrough

Following the release of artificial intelligence (AI) mathematical research breakthroughs, Ethereum Foundation researcher Justin Drake has urged the cryptocurrency industry to prepare for 'bunker mode,' a strategy involving gradually moving assets to new addresses where public keys have not been exposed. OpenAI publicly released 722 mathematical papers generated by internal models on October 6, grouped into 372 result sets, having presented approximately 4,000 problems during evaluation and using an average of about three hours of ChatGPT Pro inference computation per result. The papers covered number theory, geometry, combinatorics, and theoretical computer science.

Drake warned on October 7 that AI could threaten the security of elliptic curve digital signature algorithm (ECDSA) before quantum computers do, as ECDSA is used for transaction signing in Bitcoin (BTC) and Ethereum (ETH). While OpenAI's math achievements do not mean existing cryptography can be broken soon, Drake argued that faster AI discovery of mathematical structures could reveal new attack vectors against currently secure systems. His worst-case scenario involves classical algorithms reversing public keys to private keys, which could exploit exposed public keys from transaction signatures.

'Bunker mode' refers to moving assets to addresses never used for signing transactions, reducing the risk of exposing public keys. Drake also recommended cold wallet reinforcement, public key replacement, and hash-based signature adoption for exchanges, custodians, oracles, and layer 2 security committees. However, he cautioned against rushing asset transfers.

Ethereum co-founder Vitalik Buterin agreed with using new addresses but advised caution against abrupt transitions due to potential losses from errors during the process. Buterin also noted that lattice-based cryptography, ML-DSA, and fully homomorphic encryption (FHE) could be affected if AI accelerates solving mathematical problems, emphasizing the need for re-evaluating long-term security of quantum-resistant systems. Both Drake and Buterin highlighted hash-based cryptography as an alternative, though Buterin clarified that public key encryption requires separate mathematical structures beyond hashes alone.

Korean Source

This article is an English localization of a Korean-language crypto news report. Original headline: AI 수학 성과에 저스틴 드레이크 “가상자산 업계, 벙커 모드 준비해야”