Vitalik posted a warning on X.
It wasn't about quantum computers.
It was about AI.
For the past few years, the crypto world's biggest concern about cryptographic security has been "Q-day"—the day quantum computers become powerful enough to break existing elliptic-curve cryptography (ECDSA). Most people thought that day was still a long way off.
Vitalik and Ethereum researcher Justin Drake said today: AI could get there first.
The concern has two layers:
First: ECDSA (the current signature algorithm)
ECDSA is the signature algorithm used by the vast majority of blockchain wallets today, including ETH and BTC. Drake said AI-accelerated mathematical research could break ECDSA before quantum computers become mature.
Second: Lattice cryptography
Lattice cryptography is currently considered the leading approach to "post-quantum cryptography," which is believed to be quantum-resistant. Many people thought it would be a safe haven until the quantum era arrives. Vitalik said he believes "there is a significant probability that AI mathematical research over the next two years will seriously undermine the security of lattice cryptography."
In other words: our backup plan for defending against quantum threats could be broken by AI first.
Vitalik's recommendation:
Take this seriously, but don't rush to migrate your wallet. He said botched migration operations cause more losses than hackers do. His specific recommendation is to switch to hash-based signature schemes (WOTS, SPHINCS+), whose security is based on hash functions and doesn't rely on the mathematical assumptions of lattice cryptography.
There has been no actual security incident today, no wallets have been stolen, and no unusual price movements—this is a structural warning about the future.
But structurally speaking: if AI really does weaken the foundations of today's cryptography within two years, the underlying assumptions behind blockchain security as a whole will need to be rewritten.
Have you ever thought about this?
$ETH
$BTC
#vitalik警告ai或将加速削弱密码学安全
It wasn't about quantum computers.
It was about AI.
For the past few years, the crypto world's biggest concern about cryptographic security has been "Q-day"—the day quantum computers become powerful enough to break existing elliptic-curve cryptography (ECDSA). Most people thought that day was still a long way off.
Vitalik and Ethereum researcher Justin Drake said today: AI could get there first.
The concern has two layers:
First: ECDSA (the current signature algorithm)
ECDSA is the signature algorithm used by the vast majority of blockchain wallets today, including ETH and BTC. Drake said AI-accelerated mathematical research could break ECDSA before quantum computers become mature.
Second: Lattice cryptography
Lattice cryptography is currently considered the leading approach to "post-quantum cryptography," which is believed to be quantum-resistant. Many people thought it would be a safe haven until the quantum era arrives. Vitalik said he believes "there is a significant probability that AI mathematical research over the next two years will seriously undermine the security of lattice cryptography."
In other words: our backup plan for defending against quantum threats could be broken by AI first.
Vitalik's recommendation:
Take this seriously, but don't rush to migrate your wallet. He said botched migration operations cause more losses than hackers do. His specific recommendation is to switch to hash-based signature schemes (WOTS, SPHINCS+), whose security is based on hash functions and doesn't rely on the mathematical assumptions of lattice cryptography.
There has been no actual security incident today, no wallets have been stolen, and no unusual price movements—this is a structural warning about the future.
But structurally speaking: if AI really does weaken the foundations of today's cryptography within two years, the underlying assumptions behind blockchain security as a whole will need to be rewritten.
Have you ever thought about this?
$ETH
$BTC
#vitalik警告ai或将加速削弱密码学安全