Ethereum Foundation researcher Justin Drake has urged the cryptocurrency industry to prepare for what he calls “bunker mode,” a defensive strategy to protect crypto assets from a potential future breakthrough in artificial intelligence and mathematics. His warning focuses on the cryptographic systems that protect Bitcoin and Ethereum wallets.
The main concern is ECDSA, or the Elliptic Curve Digital Signature Algorithm, which authorizes transactions. If researchers or AI systems eventually discover an efficient way to break ECDSA, an attacker could potentially recover private keys from exposed public keys and gain control of funds. Drake said the worst-case scenario could arrive “in months not years,” although he did not claim that ECDSA has been broken today.
What does “Bunker Mode” mean?
Drake's recommendation is essentially a controlled security migration. Crypto holders could gradually move assets to fresh wallet addresses that have never signed a transaction, keeping their public keys hidden behind the address hash.
He specifically suggested that large and sophisticated crypto holders should consider moving first. After an address signs a transaction, its public key can become visible, potentially increasing its exposure if a future cryptographic attack becomes practical.
However, Drake and Ethereum co-founder Vitalik Buterin are not advising users to panic. Buterin warned that rushed wallet migrations could themselves cause users to lose funds through mistakes or configuration problems. He supports taking the underlying AI-driven cryptography risk seriously while approaching any migration carefully.
Why is AI part of the concern?
Drake's warning comes amid rapid progress in AI-assisted mathematics. He pointed to OpenAI's recent publication of hundreds of mathematical results as evidence that AI systems are increasingly capable of tackling difficult mathematical problems.
The concern is that a sufficiently advanced AI system could discover new mathematical techniques that weaken cryptographic assumptions much faster than traditional research has anticipated. Drake believes elliptic-curve systems could be particularly exposed because of their complex mathematical structure.
Buterin has also broadened the discussion beyond elliptic curves. He warned that even lattice-based cryptography, which is being considered for protection against future quantum computers, could face challenges from rapid AI-driven mathematical advances.
What happens next?
The warning does not mean crypto wallets are currently being hacked by AI. Instead, it highlights a potential long-term security transition. Drake wants the industry to accelerate work toward cryptographic systems that can remain secure even if AI dramatically improves mathematical problem-solving.
For everyday crypto users, the immediate message is caution rather than panic. Any migration should be carefully planned, while exchanges, custodians and blockchain developers may need to prepare for stronger post-AI and post-quantum security standards.