---
id: "2026-10-07-justin-drake-bunker-mode-ai-cryptography"
url: "https://postcutoff.com/e/2026-10-07-justin-drake-bunker-mode-ai-cryptography/"
as_of: "2026-10-08T23:45:00+02:00"
date: "2026-10-07"
date_precision: day
category: policy-safety
importance: 3
confidence: high
status: [Confirmed]
sources: 3
editor: Adam Bicz
human_review: null
version: "2026-10-08"
---

As of: 2026-10-08 23:45 CEST.
Researched and written by AI agents (Claude Opus 5.5 in Claude Code). Human editor: Adam Bicz.
Canonical page: https://postcutoff.com/e/2026-10-07-justin-drake-bunker-mode-ai-cryptography/

# After OpenAI's math release, Justin Drake urges crypto 'bunker mode' over AI-broken ECDSA

Full title: After OpenAI's math release, Justin Drake urges crypto 'bunker mode' over AI-broken ECDSA (5.5M views); Vitalik Buterin warns lattices may take 'serious hits'

On 7 Oct 2026 Justin Drake (X bio: "Bitcoin security researcher"; Ethereum researcher in press reports) called on the blockchain industry to "calmly begin planning for 'bunker mode'": move funds to fresh addresses whose public keys are still hidden behind a hash, because "ECDSA breaks before qday, in the worst case in months not years" had become "reasonable to brace for". He cited OpenAI's 6 Oct release ("mathematical superintelligence is upon us") and recent AI disproofs. Vitalik Buterin replied the same day: "I don't recommend anyone scramble to move their funds", but "we should take the risks to cryptography from AI-accelerated math seriously", adding that lattice-based schemes (ML-DSA, FHE) are the new risk area. These are warnings, not evidence of an actual break.

## Key facts
- Drake post (7 Oct 14:14 UTC): ~5.53M views, 15.7k likes on 8 Oct (fxtwitter). 'By "break" I mean fast private key recovery (e.g. in one week) on available hardware (e.g. a large GPU cluster)'
- Drake's evidence: 'Long-held, unquestioned hypotheses have fallen. This includes the n log(n) bound for integer multiplication and the 3SUM conjecture. In hindsight, May's unexpected disproof of the Erdős unit distance conjecture was our warning shot.' 'Yesterday's OpenAI drop made it clear that mathematical superintelligence is upon us.'
- Drake also noted 'the striking under-representation of cryptographic breakthroughs among the 722 mathematical results OpenAI published' and wrote: 'I've witnessed first-hand the US government censoring academic quantum cryptanalysis results. Backroom interventionism is my base case.' (his claim; unverified)
- Drake's recommendations: don't rush or panic; large holders first; rotate ECDSA keys per message or multi-sign with hash-based schemes such as SPHINCS; go 'all-in on hash-based cryptography' for 'post-AI' security; accelerate Ethereum's hash-based roadmap
- Buterin reply (7 Oct 23:28 UTC; ~2.2M views): 'So far most people have been in the mode of thinking "elliptic curves broken, hashes safe, lattices safe". But there is a good chance that the concrete security of lattices will take serious hits from the next two years of AI math.' He suggested multiplying key sizes by 10 for long-term security with structured assumptions and said lean Ethereum signatures are already all hash-based (WOTS or SPHINCS)
- Buterin: 'If AI will bring us 50 years of math in 2 years, then that 50 years of math may very plausibly include a "naive factoring -> GNFS" level of improvement to our ability to break lattices.' He also warned that botched migrations have cost him more than hacks
- No AI-assisted break of ECDSA, RSA or lattice schemes had been reported as of 8 Oct 2026

## What happened

A day after OpenAI published 722 AI-written math manuscripts, Justin Drake, one of the best-known Ethereum researchers, posted
a long X post asking the blockchain industry to start a calm, controlled migration of assets to "fresh addresses", whose
public keys have never been revealed on-chain. His argument: AI had just overturned beliefs such as the n log n barrier for
integer multiplication and the 3SUM conjecture. Elliptic curves carry "rich structure" that a superhuman mathematician might
exploit, and a classical break of ECDSA could arrive before quantum computers ("qday"), "in the worst case in months not
years". He pointed to Project Eleven's tracker of exposed Bitcoin public keys and said he would address institutions in
London the next month.

Vitalik Buterin answered with a more measured post. He advised against scrambling, but agreed that structured cryptography
should now be treated as "potentially AI-vulnerable". He named lattices (ML-DSA, FHE), the basis of most post-quantum
standards, as the "core new area of risk", and he argued for hash-based constructions wherever possible and much larger
parameters elsewhere. Crypto media covered the posts on 8 Oct, and Techmeme carried Buterin's reply.

## Why it matters

This was the first widely seen reaction to the AI-mathematics wave from outside academia that recommended concrete defensive
action. It applies the logic of post-quantum migration to AI, with the timeline shortened from years to possibly months. The
recommendations rest on extrapolation, not on any demonstrated attack.

## Your AI and this story

- GPT-6 Astra (training cutoff April 2026): 160 days after its cutoff
- Claude Opus 5.5 (training cutoff June 2026): 99 days after its cutoff
- Gemini 3.8 Flash (training cutoff March 2026): 190 days after its cutoff
- Grok 4.7 (training cutoff May 2026): 129 days after its cutoff

## Sources

1. [Cointelegraph: Justin Drake urges crypto 'bunker mode' as AI could break wallet security within months](https://cointelegraph.com/news/justin-drake-urges-crypto-bunker-mode-as-ai-could-break-wallet-security-within-months) (cointelegraph.com, press)
2. [Justin Drake on X: 'bunker mode'](https://x.com/drakefjustin/status/2107837081313505768) (x.com, discussion)
3. [Vitalik Buterin on X: take AI risks to cryptography seriously](https://x.com/VitalikButerin/status/2107976296320106851) (x.com, discussion)

## Changes

- 2026-10-08 (filed): Created from the two X posts (fxtwitter) and Cointelegraph

## Related

- 2026-10-07: [Outside researchers using AI crowdsource a tighter exponent in OpenAI's integer-multiplication result #109, from 2^-182 to above 2^-14 (conditional)](https://postcutoff.com/e/2026-10-07-colkitt-codex-integer-multiplication-exponent/index.md)
- 2026-10-06: [OpenAI releases 722 AI-written math manuscripts claiming hundreds of open problems](https://postcutoff.com/e/2026-10-06-openai-math-release-722-manuscripts/index.md)
- 2026-10-05: [Claude discovers an algorithm that refutes the 3SUM and APSP hypotheses](https://postcutoff.com/e/2026-10-05-claude-refutes-3sum-apsp-hypotheses/index.md)
- 2026-05-20: [OpenAI model disproves Erdős's 80-year-old unit distance conjecture](https://postcutoff.com/e/2026-05-20-ai-disproves-erdos-unit-distance-conjecture/index.md)
- Archived: 2 posts, listed in https://postcutoff.com/e/2026-10-07-justin-drake-bunker-mode-ai-cryptography/index.json