Ethereum Drafts EIP-8394: A Quantum-Safe Fix for 96-Byte Keys

On August 24, researcher Kevaundray Wedderburn filed a draft proposal to rebuild Ethereum’s validator deposit contract for a post-quantum future. Co-authored with Thomas Coratger and Tom Wambsgans, the draft introduces variable-length keys and a switch that would permanently retire BLS signatures, the scheme every validator uses today to join the network. An editor has suggested the number EIP-8394 for it, though it isn’t yet published on Ethereum’s official EIP site. The filing lands five months after a Google Quantum AI paper cut the estimated cost of breaking Ethereum’s cryptography by 20 times.

A New Deposit Contract Takes Shape

Ethereum’s deposit contract was built around one fixed structure: a 48-byte BLS public key and a 96-byte BLS signature. Wedderburn’s draft replaces that with variable-length public keys, signatures and credential metadata, plus a scheme identifier that records which cryptographic system each validator uses. Scheme 0 stays reserved for existing BLS deposits.

The rollout is staged, not immediate. The contract would launch with BLS deposits still active while new schemes register alongside it. A later, separate protocol decision could then permanently disable new Scheme 0 deposits — a switch that, once flipped, cannot be reversed under the current design. Existing BLS validators wouldn’t disappear when that happens; handling their exits, credential changes and eventual migration is a harder problem the draft doesn’t attempt to solve yet.

The design leans on EIP-7685, the framework Ethereum already uses to move deposits, withdrawals and validator consolidations from the execution layer to the consensus layer. Wedderburn has said the proposal still needs to reconcile with three existing validator EIPs — 7002, 7251 and 8282 — and that the topic surfaced during an All Core Developers Execution call.

Why BLS Signatures Are at Risk

BLS signatures let Ethereum compress votes from hundreds of thousands of validators into one compact proof, relying on elliptic curve pairings for that efficiency. Those same pairings are what a sufficiently powerful quantum computer running Shor’s algorithm could break. No such machine exists yet, but the timeline compressed on March 31, 2026, when Google Quantum AI, working with Ethereum Foundation researcher Justin Drake and Stanford’s Dan Boneh, published a paper showing that breaking 256-bit elliptic curve cryptography could take fewer than 500,000 physical qubits — down from roughly 9 million in prior estimates, a 20-fold drop.

Ethereum’s proposed fix, leanXMSS, is a hash-based signature scheme that Shor’s algorithm cannot break. The tradeoff is size: a leanXMSS signature runs about 3,000 bytes, roughly 30 times larger than a BLS signature, which is exactly why the deposit contract needs restructuring before the new keys can fit.

Regulators are moving on a parallel clock. NIST’s draft report IR 8547 sets 2030 as the deadline to begin deprecating ECDSA, RSA and EdDSA, with full disallowance after 2035.

The Road to a 2029 Deadline

This deposit-contract draft is one piece of a broader plan Vitalik Buterin laid out in a February 2026 roadmap, which splits Ethereum’s post-quantum work into four tracks:

• Validator BLS signatures tied to deposit contracts
• KZG commitments used in data availability
• ECDSA signatures protecting ordinary accounts
• Zero-knowledge proof systems used by rollups

Account-level protection runs on a separate track. EIP-8141 introduces native account abstraction so individual accounts can choose their own signature scheme, and it’s under consideration for the Hegotá upgrade.

The Ethereum Foundation stood up a dedicated post-quantum team in January 2026, led by Thomas Coratger, and backed it with a $1 million Poseidon Prize for hash-function research on top of an existing $1 million Proximity Prize. Justin Drake called the effort a “top strategic priority.” Full protocol-level readiness is targeted for around 2029 — a planning goal, not a locked deadline.

Conclusion

Wedderburn’s draft doesn’t change anything on Ethereum mainnet today. It’s an early-stage pull request, still failing some automated checks, still waiting on a formal editor assignment, and still missing rules for migrating existing validators once BLS retires. But it does something more important: it gives Ethereum’s deposit contract room to hold keys it was never built to store. With NIST’s 2030 deadline approaching and quantum-resource estimates shrinking fast, that flexibility is the groundwork the rest of the migration depends on.