Scenarios for Post-Quantum Migrations

2026 Feb 11 See all posts


Scenarios for Post-Quantum Migrations

A meme in the cryptographic domain is “Cryptography is easy, Key Management is hard”. While current efforts to bring Ethereum to the Post-Quantum Era are underway, deployment will still take years. This post does not venture to make any bet or commitment on which scenario is most probable for which deadline. But first, let’s examine how to handle a risk.

Basically, risk resource allocation follows:


risk_materialization_cost × risk_probability = resource_allocation

Looking at today’s risk management in the Ethereum ecosystem, it seems to be handled more as a binary ‘0/1’ problem, either quantum computers are a threat or they aren’t. There’s no widely accepted model to extrapolate quantum computing progress, making probabilistic planning difficult. What we can do is prepare for multiple scenarios. First let’s have a look at the exposure:

Category Description Estimated Value Vulnerability Level Notes
ETH in EOAs (exposed pubkey) Any EOA that has ever sent a transaction $140-175B 🔴 Critical ~40-50M ETH, pubkey on-chain
Staked ETH (validators) 32 ETH × ~1M validators ~$110B 🔴 Critical BLS keys compromised, consensus at risk
DeFi Protocol Treasuries DAO treasuries in EOAs/multisigs $5-15B 🔴 Critical Many use EOA signers
Bridge Hot Wallets Cross-chain bridge reserves $8-12B 🔴 Critical Multisigs with EOA keys
CEX Hot Wallets Exchange operational wallets $15-25B 🔴 Critical High-value targets
CEX Cold Storage Offline/MPC storage $50-100B 🟡 Medium May have airgapped keys
ERC-20 Tokens in EOAs All tokens in vulnerable addresses $80-120B 🔴 Critical Multiplier effect
Centralized Stablecoins (USDT/USDC/etc.) Centralized $120B+ 🟢 Low (USDT/USDC) Centralized can freeze
NFTs (major collections) BAYC, Punks, Art Blocks, etc. $2-5B 🔴 Critical Sell them now, they are already dead
L2 Sequencer Keys Rollup operational keys N/A (control risk) 🔴 Critical Can censor/manipulate L2
ENS Names Valuable domain names $0.5-1B 🔴 Critical In EOAs

This post explores various scenarios of key migration in the face of the Quantum Computing Nuke.

pektra

(Now pick your difficulty level)


PQBeat Readiness Stages

Before diving into scenarios, let’s define the PQBeat stages that measure an ecosystem’s post-quantum readiness. These stages are tracked at pqbeat.zknox.com.

L1/L2 Chain Stages

Stage Criteria Technical Requirements
Stage 0 No PQ capability deployed ECDSA only
Stage 1 PQ accounts usable in production ERC-4337 / EIP-7702 + PQ signature verification (via smart contract)
Stage 2 Native PQ accounts in production EIP-7701 + precompiles (ML-DSA/Falcon/XMSS/...)
Stage 3 Migration path operational ZK-recovery + freeze mechanism
Stage 4 Post-Quantum Consensus Layer Previous stages + BLS replaced by aggregatable PQ signatures

Wallets & Protocols

Wallets and protocols inherit the stage of their deployment chain when they support PQ smart accounts:

Stage Criteria
Stage 0 No PQ support (ECDSA only, EOA-gated)
Stage N PQ smart accounts supported on a Stage N chain

The Current Efforts

The Ethereum community is actively working on post-quantum migration strategies. The emerging consensus points toward a layered approach: aggregating PQ signatures into Post-Quantum SNARKs (specifically STARK or WHIR-based proofs) to manage the significant size overhead of PQ signatures on-chain.

Several concrete implementations already exist. Post-Quantum Smart Accounts based on ERC-4337 have been developed, notably hash-based schemes like XMSS with ZK-friendly Poseidon hashing and ZKNOX’s implementation supporting Falcon, Dilithium (ML-DSA). These allow users to transition to quantum-resistant security today, without waiting for protocol-level changes.

The key derivation architecture would extend BIP32/BIP39 with new derivation paths for each PQ algorithm, using the same master seed to derive both legacy secp256k1 keys and new PQ keys (Falcon at m/44'/9002'/..., Dilithium at m/44'/9003'/..., etc.). The following diagram illustrates the foresight mechanism, adapted to the ZKNOX Post Quantum key derivation scheme.

pektra

(ZKNOX Key Derivation Standard for PQC)

The concept of ZK-based key migration for frozen accounts has been described by Vitalik Buterin. The mechanism is elegant: users don’t need to pre-register anything. They simply keep their BIP39 seed phrase. After a freeze of legacy EOAs, recovery works by submitting a ZK proof demonstrating knowledge of the master seed that generated both the legacy ETH address AND a new PQ public key. The proof links the old identity to the new key without revealing the secret. To date, however, no production implementation of this recovery mechanism exists.

The following diagram illustrates the circuit related to the ZK-BIP39 proof.

pektra

(ZK Based Key Migration)

Piece of Cake (Easy Mode): The Smooth Migration

Required Stage Impact Losses
Stage 4 reached before CRQC ✅ None $0

pektra

(The world is cute when everything goes fine)

In this optimistic scenario, the cryptographic community provides sufficient warning — perhaps a decade or more — before cryptographically relevant quantum computers (CRQCs) emerge. Users, DeFi protocols, and infrastructure providers have ample time to migrate.

What happens:

The timeline is comfortable: perhaps 2-3 years for protocol changes, another 2-3 years for ecosystem adoption, all completed well before any quantum threat materializes.

User experience: Seamless. A wallet notification prompts upgrade, user confirms, and their account is now quantum-resistant, still controlled by the same seed phrase.

Stage impact on actors:

Actor Starting Stage End Stage Migration Path
Ethereum L1 Stage 0 Stage 4 Precompiles → ZK-recovery → CL migration
Major L2s Stage 0 Stage 3+ Inherit L1 precompiles, add PQ sequencer keys
Ledger/Trezor Stage 0 Stage 2+ Firmware update with PQ key derivation
Uniswap/Aave Stage 0 Stage 2 Accept PQ smart accounts as signers
Bridges Stage 0 Stage 2 PQ multisig implementations

This is the scenario everyone hopes for. It requires no emergency measures, no frozen funds, no losses.


Let’s Rock (Normal Mode): Quantum Computing Accelerates

Required Stage Impact Estimated Losses
Stage 3 reached before CRQC ⚠️ Moderate $1-5B (dormant wallets)

pektra

(Let’s be cautious)

In this scenario, quantum computing progress outpaces initial estimates, but the community responds in time. Strong signals, perhaps a demonstration of Shor’s algorithm on increasingly large key sizes, trigger coordinated action.

What happens:

Stage requirements at freeze time:

Actor Required Stage Status
Ethereum L1 Stage 3 ZK-recovery + freeze operational
Wallets Stage 2+ Can generate ZK proofs from seed
DeFi protocols Stage 1+ Accept PQ smart accounts
Users (active) Stage 1+ Already migrated
Users (dormant) Stage 0 Must use ZK-recovery

Post-freeze recovery:

The ZK recovery mechanism activates. Users with frozen accounts can recover by generating a WHIR-aggregated proof locally (using their seed phrase) and submitting it alongside their new PQ public key. The proof demonstrates: “I know the seed that generated this frozen address, and here’s my new PQ key derived from the same seed.”

Wallet vendor support: Hardware and software wallet vendors ship “recovery mode” firmware (requires Stage 2 capability). The user connects their device, enters their seed phrase (or it’s already stored securely), and the wallet generates the witness data for the ZK prover. The UX could be as simple as: “Your account was frozen for quantum security. Click here to recover with your new quantum-resistant key.”

Economic assumption: The cost of mounting a Shor attack (requiring a fault-tolerant quantum computer with thousands of logical qubits) exceeds the value of most individual wallets. Attackers would target the largest honeypots first, giving smaller holders time to recover.

Losses: Minimal. Some long-lost wallets (Satoshi-era ETH, forgotten accounts stuck at Stage 0) remain frozen forever, their owners either lost their seeds or are unreachable. But active economic participants recover successfully.


Come Get Some Losses (Hard Mode): Wow, Didn’t See It That Fast

Ecosystem Stage Impact Estimated Losses
Stage 1-2 (incomplete Stage 3) 🔴 Severe $20-40B+

pektra

(That was close)

In this scenario, the transition is underway but incomplete when quantum capabilities arrive. Stage 3 hasn’t been reached — no operational freeze/recovery mechanism. Some major actors are caught exposed.

Ecosystem state at CRQC arrival:

Actor Stage Vulnerability
Ethereum L1 Stage 2 Precompiles deployed, no freeze mechanism
Early adopters Stage 1-2 Safe (migrated to PQ accounts)
Most users Stage 0 🔴 Exposed pubkeys vulnerable
DeFi protocols Stage 0-1 Mixed — some accept PQ, most don’t
Validators Stage 0 🔴 BLS keys compromised

What happens: - A cryptographically relevant quantum computer is demonstrated or credibly rumored - Panic ensues; sophisticated attackers (nation-states, well-funded groups) begin targeting high-value addresses - The public keys of any address that has ever sent a transaction are now vulnerable — attackers can derive private keys from exposed public keys

Immediate damage assessment:

Category Estimated Exposure Stage Required to Avoid
ETH in EOAs with exposed pubkeys ~70% of all EOA-held ETH Stage 1+ migration
Top 100 non-contract addresses ~$10-15B Stage 1+ migration
DeFi protocol treasuries (EOA-held) ~$2-5B Stage 1 protocol upgrade
Bridge hot wallets ~$1-3B Stage 1+ multisig upgrade
CEX hot wallets ~$5-10B Internal PQ migration

Total potential loss: $20-40B+ in immediate exposure, assuming current (2024-2025) valuations.

Emergency response without Stage 3: - Centralized stablecoins freeze: Tether (USDT) and Circle (USDC) have admin keys capable of blacklisting addresses. They freeze all movement pending resolution. This contains ~$100B+ in value but requires trusting centralized actors. - CEX circuit breakers: Major exchanges halt withdrawals and deposits. - Emergency Stage 3 deployment: Rushed implementation of freeze mechanism — high risk of bugs - Rollback debate: The community faces an agonizing choice, roll back the chain to before the attacks, losing potentially days or weeks of legitimate transactions, or accept the losses and move forward.

The rollback calculus: A rollback is technically possible but socially catastrophic. It would:

The 2016 DAO hack rollback worked because the community was smaller and the damage was contained. A quantum attack could be far more distributed and harder to cleanly separate from legitimate activity.

Likely outcome: A messy compromise, partial freezes, emergency hard fork to disable secp256k1, rapid deployment of ZK recovery (rushed Stage 3), and significant but not total losses. Some whales get rekt. Trust is damaged but Ethereum survives.


Damn I’m (not) Good (Nightmare Mode): We Are So Rekt

Ecosystem Stage Impact Estimated Losses
Stage 0-1 (no preparation) 💀 Catastrophic $100B+ / Chain survival at risk

pektra

(REKT)

In this scenario, quantum computing is available tomorrow. No warning. No preparation complete. Shor’s algorithm runs on a machine that can break secp256k1 in hours or days.

Ecosystem state: Almost entirely Stage 0

Actor Stage Status
Ethereum L1 Stage 0 No precompiles, no freeze, no recovery
L2s Stage 0 ECDSA-only
Wallets Stage 0 No PQ support
DeFi Stage 0 EOA-only governance
Validators Stage 0 BLS12-381 compromised
ZKNOX users Stage 1 🟢 Only survivors with self-custody

Immediate impact:

Emergency possibilities:

1. Centralized Currency Freeze

USDT, USDC, and other centralized stablecoins can freeze all transfers via admin functions. These issuers would wait until a recovery mechanism is designed, then selectively unfreeze based on proof of legitimate ownership.

2. Naive Consensus Fallback (Stage 4 emergency improvisation)

A hastily-deployed consensus mechanism where each validator signs using a PQ signature (e.g., Dilithium) before BLS aggregation could theoretically work, but the performance impact would be severe:

Current BLS aggregation: - Signature size: 96 bytes (aggregated from all attesters)

Naive PQ replacement (FALCON512):

This is obviously unworkable for normal operation. The network would grind to a halt or require massive validator set reduction (e.g., only 1,000 validators, re-introducing centralization), or some recursion.

A more realistic emergency consensus:

3. ERC-4337 Account Abstraction Refuge (Stage 1 lifeboats)

Users with existing 4337 smart accounts using PQ verification (like ZKNOX accounts) would be the only ones with secure, self-custodial access. These Stage 1 accounts become lifeboats. However, if the underlying L1 consensus is compromised, even these accounts are only as secure as the chain they’re on.

4. L2 Evacuation

If L2s can quickly deploy PQ verification (emergency Stage 1) and their sequencers aren’t compromised, they might serve as temporary safe havens. Assets could be bridged to L2s that implement emergency PQ measures faster than L1 can hard fork.

Realistic outcome: Catastrophic even if not terminal. Ethereum would survive in a very diminished, emergency form. Massive value destruction occurs. Recovery takes months or years. The “immutable ledger” narrative is permanently damaged, but the technology eventually rebuilds with proper quantum resistance.


Summary: Scenarios vs. Required Stages

Scenario Difficulty Required Stage Time to Prepare Expected Losses
Piece of Cake Easy Stage 4 5-10 years $0
Let’s Rock Normal Stage 3 2-5 years $1-5B (dormant)
Come Get Some Hard Stage 2 (incomplete 3) 1-2 years $20-40B
Damn I’m Good Nightmare Stage 0-1 0 (too late) $100B+

stages

(Where are we? Where do we need to be?)


Conclusion

pektra

(We need a red button .)

A kill switch is required, not to use, hopefully never to use, but to have ready as a mitigation against the nightmare scenario. In PQBeat terms: we need Stage 3 readiness as insurance, even if we hope to reach Stage 4 gracefully.

This means:

  1. Stage 3 minimum: Implemented ZK recovery circuits, tested and audited, ready for emergency deployment

  2. WHIR aggregation infrastructure that can handle mass recovery operations

  3. Wallet vendor integration (Stage 2) so recovery UX is ready, not improvised

  4. Emergency consensus specifications (Stage 4 fallback) that trade decentralization for survival

  5. Coordination mechanisms between major stakeholders (client teams, large stakers, centralized issuers) for rapid response

The cost of building this infrastructure is trivial compared to the potential losses. Even if the probability of a nightmare scenario is low, the expected value calculation clearly favors preparation:


$4Mpreparation ≪ 0.1% × $40Bloss = $400Mexpected_loss

The Nuke may never come, or may come so slowly we barely notice the transition. But if it comes fast, we need to be ready.

Track progress at pqbeat.zknox.com. Keep your seed phrase safe. It may be your only recovery path.

broken

(Quantum Apocalypse Clock is ticking)

References

Post-Quantum Readiness in EdDSA Chains

How to hard-fork to save most users’ funds in a quantum emergency

PQBeat - Post-Quantum Readiness Tracker


Reach us

🔐 Practical security on the whole chain.

Github | Website | Twitter | Blog | Contact Info

Found typo, or want to improve the note ? Our blog is open to PRs.

ZKNOX — Post-Quantum Cryptography for Ethereum