Vitalik Buterin says there’s a ‘good chance’ AI could break lattice cryptography within two years

For years, the big scary threat to crypto security has been the quantum computer. Vitalik Buterin now says it may have company.

On October 7, 2026, the Ethereum co-founder warned that AI-driven advances in mathematics could weaken lattice-based cryptography. He said there is a “good chance” that lattice security could be compromised within the next two years.

What Buterin actually said

Buterin’s warning centers on lattice-based cryptography. This family of math underpins several systems built specifically to survive the quantum era.

He named two systems in particular that could be significantly undermined: ML-DSA and Fully Homomorphic Encryption (FHE). ML-DSA is a digital signature scheme designed to resist quantum attacks. FHE is a technique that lets computers run calculations on encrypted data without ever decrypting it.

Buterin also flagged a growing risk to ECDSA signatures, which sit alongside the quantum threat. ECDSA is the signature system that crypto wallets commonly use to authorize transactions. He highlighted that the danger rises as public keys become exposed on-chain.

Don’t panic, but do switch locks

Despite the alarming timeline, Buterin urged people not to make panic-driven moves with their assets.

First, he said hash-based signatures should be preferred over lattice schemes wherever that’s feasible. Second, he advised using fresh addresses that have never executed a transaction, as an address that has never sent anything hasn’t yet revealed its public key, which shrinks the attack surface.

Justin Drake, an Ethereum researcher, made a similar point on the same day. Drake emphasized that recent breakthroughs in AI-driven mathematics could produce vulnerabilities in ECDSA signatures much sooner than previously anticipated. He called for faster preparation for that possibility.

Why Ethereum was already moving

Buterin’s comments line up with a direction Ethereum had already chosen. The network’s Lean roadmap has shifted toward hash-based signature schemes such as WOTS and SPHINCS+.

The goal is to reduce dependence on cryptographic methods that look vulnerable to quantum attacks, AI-driven attacks, or both.

Bigger than blockchains

The risks discussed here are not limited to crypto. The conversation explicitly extends to secure websites, messaging apps, and VPNs.

What this means for holders and builders

No immediate market impact was linked to the warnings from Buterin or Drake. For individual holders, the clearest takeaway is Buterin’s guidance on address hygiene. Using fresh addresses that have never transacted limits how much information an attacker can work with.

For builders, the message is about design choices. Projects leaning on lattice-based schemes for signatures or encryption may want to revisit those decisions. Where hash-based alternatives are workable, Buterin is clearly signaling a preference for them.

Buterin framed his concern as a probability, not a certainty, and the two-year window is a forecast. Any concrete AI-assisted result against lattice schemes or ECDSA would turn this from a warning into an emergency.