The Ledger
Preprint proposes one-transaction proofs for Ethereum quantum migration
A research paper proposes a seed-proof system to let Ethereum account owners move to post-quantum keys, including after a fraudulent enrollment. Its authors report that a proof of about 190 KB could fit in one transaction and estimate migration times in a ten-node simulation.

A proposed Ethereum migration would let account owners prove control of a wallet’s original seed in one transaction, including to correct a fraudulent post-quantum key registration. The arXiv preprint, submitted on Nov. 19, describes a plan for moving accounts and validators away from cryptographic keys that could be vulnerable to a sufficiently capable quantum computer.
The authors estimate that accounts exposed by their own transactions hold 54.6% of execution-layer ETH. The paper argues that a migration based only on existing signatures could be vulnerable if an attacker could extract account keys before the network’s transition: an attacker might register a replacement key in an owner’s name.
A correction window for enrolled keys
The proposed migration has two stages. Owners could first register post-quantum keys using their existing signatures, without a zero-knowledge proof, before ECDSA is retired at a scheduled time. The paper says a seed proof could later correct an enrollment until a common deadline.
An account that was not enrolled by the cutoff would be frozen, not destroyed, and could be recovered with a seed proof without a deadline. The paper says the proof would demonstrate knowledge of the wallet seed used to derive the account’s key. It would bind the replacement key, a nonce and a fee cap, which the authors say would prevent a sponsor or a block builder from substituting a key, replaying a proof or charging above the cap.
Validators and proof size
The design would switch validators to hash-based keys first, at a finalized checkpoint. The authors describe this step as dependent on a stated bound on an attacker’s voting power. They say it is intended to let account recoveries reach finality after the transition.
For account recovery, the paper describes a three-layer, recursive hash-based STARK proof covering the derivation of an Ethereum account from a BIP-39 seed. The authors report that the proof is about 190 KB, fits in one transaction, verifies in 38 ms and takes a median of 111 seconds to generate on a laptop. They also say four owners’ proofs can be combined into one batch proof.
Simulation estimates migration time
The authors tested the design in a ten-node simulation using mainnet-derived accounts and fee demand. For 4.45 million accounts holding at least 0.1 ETH, they estimate migration would take 29.4 days if all owners enrolled before ECDSA was retired. If all instead recovered afterward in batches of four, their estimate was 235 days.
The paper also reports a median public-mempool wait of 19.1 minutes for an enrollment during mass migration. It compares that delay with published estimates for a fast quantum computer to extract a key, and says owners with a seed and an unexposed key should use a seed proof rather than enrollment. These are findings and recommendations presented by the paper; it describes a proposed migration, not an adopted Ethereum upgrade.
Sources
- Original source · Preprint Correctable Enrollment and One-Transaction Seed Proofs for Ethereum's Post-Quantum Transition arXiv:2611.25663
Related stories
- Preprint proposes fee-memory rule for Layer 2 posting costs
- Researchers measure a hash-based verifier for Bitcoin bridges
- Preprint renames LinkDID as Braid, an identity design that links identifiers across apps
- Researchers update Angelfish consensus preprint
- Lightning force-closes cost 2–6 times more in fees, preprint finds