Executive Summary
On April 15, 2026, Google Quantum AI published a whitepaper demonstrating that breaking the elliptic curve cryptography securing Bitcoin and Ethereum requires roughly 20 times fewer quantum resources than previously estimated. The required threshold has been revised downward to under 500,000 physical qubits, with as few as 1,200–1,450 high-quality logical qubits needed for a practical attack.
No cryptographically relevant quantum computer (CRQC) exists today — current demonstrated capability stands at 24–28 logical qubits, orders of magnitude below the attack threshold. However, the compressed timeline forces institutional risk models to treat Q-Day as a medium-term planning horizon rather than a generational concern, with Google itself setting a 2029 internal migration deadline.
A May 2026 Glassnode analysis found approximately 6.04 million BTC (representing 30.2% of circulating supply) in wallets with publicly visible keys, making them theoretically targetable in a post-quantum world. Both Bitcoin and Ethereum have active post-quantum cryptography (PQC) roadmaps, though they differ fundamentally in governance, architecture, and urgency.
Key Findings
- The quantum threat to ECDSA/Schnorr signatures is real and the timeline has compressed — but not yet imminent.
- 30–33% of Bitcoin supply has exposed public keys, including ~6.9M BTC per Google’s own estimate.
- Bitcoin’s 2021 Taproot upgrade inadvertently widened quantum exposure by placing public keys on-chain by default.
- NIST PQC standards (ML-DSA, SLH-DSA) are technically feasible but introduce severe on-chain size constraints.
- SHA-256 (Bitcoin mining) remains secure; the threat is specific to signature schemes, not proof-of-work.
1. The Quantum Threat to Blockchain Cryptography
Widely deployed public-key cryptography — specifically the Elliptic Curve Digital Signature Algorithm (ECDSA) and Schnorr signatures on the secp256k1 curve — is built on the mathematical hardness of the Elliptic Curve Discrete Logarithm Problem (ECDLP). Shor’s algorithm, running on sufficiently powerful quantum hardware, can solve the ECDLP in polynomial time, reconstructing a private key from a public key. This renders every wallet whose public key is visible on-chain theoretically vulnerable.
1.1 The Harvest Now, Decrypt Later Mode
Adversaries do not need a CRQC to begin exploiting blockchains today. Under the Harvest Now, Decrypt Later (HNDL) model, on-chain public key data is permanently recorded and available for future decryption. Unlike encrypted communications that can be re-secured, blockchain public keys that have ever been exposed cannot be retroactively hidden.
1.2 Threat Classification by Attack Profil
Quantum attacks on blockchain assets fall into two distinct profiles with different urgency and mitigation requirements:
- At-Rest Attacks (Exposed Public Keys): The most severe and immediate category. On blockchains, a public key is permanently recorded when a transaction is broadcast. Vulnerable outputs include early Satoshi-era Pay-to-Public-Key (P2PK) addresses, modern Taproot (P2TR) keypath spends, and any address where keys have been reused. A CRQC could derive private keys from these at leisure, with no time pressure.
- On-Spend Attacks (In-Transit Transactions): A narrower window attack targeting transactions in the mempool. An attacker must derive the private key and broadcast a competing transaction before block confirmation — roughly 10 minutes on Bitcoin and 12 seconds on Ethereum. This requires far more powerful quantum hardware and real-time operation.
1.3 The Google Quantum AI Whitepaper
Google Quantum AI whitepaper titled « Securing Elliptic Curve Cryptocurrencies Against Quantum Vulnerabilities » presented the most aggressive resource estimates to date:
- Breaking the 256-bit elliptic curve cryptography underlying Bitcoin and Ethereum would require fewer than 500,000 physical qubits — a reduction of approximately 20× compared to previous estimates exceeding 10 million qubits.
- Equivalently, as few as 1,200–1,450 high-quality logical qubits could suffice for a practical attack (prior estimates placed this at 2,330 or higher).
- The paper modelled a real-time mempool attack with a 41% success rate against Bitcoin’s 10-minute block confirmation window, completing in approximately 9 minutes on a superconducting architecture.
- An alternative model (Cain et al., 2026) suggests approximately 10,000 reconfigurable atomic qubits supported by 26,000 physical qubits could complete key derivation in days using high-rate error-correcting codes.
2. Bitcoin’s Quantum Exposure: Updated Supply Analysis
Quantifying exactly how much Bitcoin is at risk requires distinguishing between structural exposure (address formats that expose public keys by design) and operational exposure (user behaviour such as key reuse). A recent Glassnode analysis provides the following mapping.
| Address Type | BTC at Risk | % of Supply | Notes |
| P2PK (Satoshi-era) | ~1.7M BTC | ~8.5% | Keys exposed by design; majority likely lost, cannot migrate |
| P2TR (Taproot) | ~4.2M BTC* | ~21% | Public key on-chain by default since Nov 2021 activation |
| P2MS (legacy multisig) | Small residual | <1% | Structurally exposes keys; largely replaced |
| Total Structural (Glassnode) | 1.92M BTC | 9.6% | Script type exposes key regardless of owner action |
| Total Operational (Glassnode) | 6.04M BTC | 30.2% | Includes reused addresses and all exposed-key formats |
| Google Estimate (March 2026) | ~6.9M BTC | ~33% | Wider estimate including Taproot keypath spends |
3. NIST Post-Quantum Cryptography Standards: Feasibility and Trade-off
In 2024, the U.S. National Institute of Standards and Technology (NIST) finalised its first set of Post-Quantum Cryptography (PQC) standards. These represent the cryptographic schemes blockchains must migrate to if they are to resist Shor’s algorithm. The standards include:
- ML-DSA (formerly CRYSTALS-Dilithium): A lattice-based digital signature scheme. The primary candidate for blockchain signature migration.
- SLH-DSA (formerly SPHINCS+): A hash-based signature scheme offering extremely conservative, well-understood security assumptions.
- FN-DSA (Falcon): A more compact lattice-based scheme under consideration for Ethereum precompiles.
3.1 Computational Performance
Benchmarking studies (Schemitt et al., October 2025) indicate PQC signature schemes introduce only minor computational overhead at Security Level 1 and can outperform ECDSA at higher security levels. On an ARM-based laptop, ML-DSA achieves signature verification of 0.14ms at Security Level 5, versus 0.88ms for ECDSA. Execution time is not the binding constraint.
3.2 The Critical Bottleneck: On-Chain Size
The binding architectural constraint is the dramatic increase in public key and signature sizes, which cascades into fundamental trade-offs across every layer of blockchain design.
|
Scheme |
Public Key |
Signature |
Size vs ECDSA |
Blockchain Impact |
|
ECDSA (secp256k1) |
32–64 bytes |
64–73 bytes |
Baseline |
Current standard; minimal block overhead |
|
Falcon-512 (FN-DSA) |
897 bytes |
666 bytes |
~10× larger sig |
Most compact lattice scheme; complex floating-point implementation |
|
ML-DSA-44 (Dilithium2) |
1,312 bytes |
2,420 bytes |
~37× larger sig |
Simpler integer arithmetic; reduces TPS, raises gas |
|
SLH-DSA-128f (SPHINCS+) |
32 bytes |
16,976 bytes |
~250× larger sig |
Hash-only security; economically prohibitive on-chain |
The practical consequences of these size increases are severe:
- Block sizes must expand significantly to accommodate PQC signatures, directly reducing transactions per second (TPS) throughput.
- Larger transactions increase network propagation latency and elevate block orphan (fork) rates.
- On-chain EVM verification costs are prohibitive without precompiles: vanilla ML-DSA-44 signature verification currently costs approximately 1.8 million gas on the EVM, compared to just 3,000 gas for ECDSA (ecrecover). Proposed precompiles (EIP-8051 for ML-DSA, EIP-8052 for Falcon-512) aim to reduce this to approximately 3,000 gas, but are not yet deployed on mainnet.
4. Bitcoin’s Post-Quantum Roadmap
The Bitcoin development community has produced two distinct but complementary proposals that together form a two-stage quantum migration roadmap. Critically, these are separate Bitcoin Improvement Proposals (BIPs) and should not be conflated.
4.1 BIP-360: Pay-to-Merkle-Root (P2MR)
BIP-360 was introduced in December 2025, as a first step in advancing Bitcoin quantum resistance. it addresses the at-rest vulnerability by eliminating the keypath spend mechanism that currently exposes public keys on-chain in Taproot outputs.
- The Problem: P2TR (Taproot) outputs include a tweaked public key in plaintext, exposing it to long-range quantum attack via Shor’s algorithm.
- The Solution: BIP-360 introduces a new SegWit version 2 output type (P2MR) that commits directly to the Merkle root of the script tree, with no public key appearing on-chain until a specific script branch executes. Addresses use the bc1z prefix with bech32m encoding.
- Preserved functionality: Lightning Network, BitVM, Ark, and time-lock contracts all remain operational through the P2MR tapscript path.
- Limitation: BIP-360 addresses long-exposure at-rest attacks. It does not protect against short-exposure on-spend attacks, where a quantum computer would need to break a signature within the transaction confirmation window. Full protection requires the future adoption of PQC signature opcodes (Dilithium is implemented in the testnet but not yet formalised in a mainnet BIP).
4.2 BIP-361: Post-Quantum Migration and Legacy Signature Sunset
BIP-361 — a separate proposal formally assigned on February 11, 2026, establishes the phased migration and legacy sunset timeline. It is currently in informational status and requires no immediate action from holders.
- Phase A — Transitional Window (Year 3 post-activation): Consensus rules reject any new outputs created with legacy ECDSA or Schnorr scripts. All newly minted or transferred coins must use quantum-resistant P2QRH or P2MR addresses.
- Phase B — Legacy Sunset (Year 5 post-activation): A hard ‘flag day’ after which all nodes reject spending transactions signed using classical ECDSA or Schnorr, effectively freezing unmigrated dormant outputs.
- Phase C — Zero-Knowledge Recovery: For users who missed the deadline, a potential recovery path using ZK-SNARK proofs of possession of the legacy BIP-39 seed phrase, allowing claims without revealing vulnerable public keys.
4.3 Live Testnet: Bitcoin Quantum v0.3.0
BTQ Technologies launched Bitcoin Quantum testnet v0.3.0 in March 2026 — the first working implementation of BIP-360. As of release:
- Full P2MR consensus with SegWit v2 outputs and bc1z address encoding (bech32m).
- All five Dilithium post-quantum signature opcodes enabled in P2MR tapscript context.
- End-to-end CLI wallet tooling covering address creation, funding, transaction construction, signing, mempool acceptance, broadcast, and confirmation.
- Over 50 miners participating; more than 100,000 blocks processed.
5. Ethereum’s Post-Quantum Roadmap: The Strawmap
In February 2026, Vitalik Buterin published Ethereum’s post-quantum roadmap — rebranded as the Strawmap — identifying four distinct cryptographic layers requiring post-quantum upgrades. The approach is more structured than Bitcoin’s, targeting core infrastructure completion by 2029 across approximately seven hard forks.
The four vulnerable layers are: (1) consensus-level BLS validator signatures; (2) KZG-based data availability commitments; (3) ECDSA account signatures (the execution layer); and (4) ZK-proof systems relying on elliptic curve pairings.
5.1 Execution Layer: Signature Agility via Native Account Abstraction
Rather than a mandatory wallet-replacement hard fork, Ethereum’s execution-layer strategy relies on account abstraction to enable voluntary, gradual migration:
- EIP-7702 (deployed in Pectra, May 2025): Allows externally owned accounts (EOAs) to temporarily execute smart contract code, unlocking transaction batching, gas sponsorship, and social recovery — a stepping stone to full account abstraction.
- EIP-8141 (Frame Transactions): Under consideration for the Hegotá hard fork (planned H2 2026). Introduces native protocol-level account abstraction, splitting transaction execution into VERIFY, SENDER, and DEFAULT stages. This allows users to swap signature schemes from ECDSA to ML-DSA or Falcon without abandoning their accounts.
- Gas precompiles: EIP-8051 (ML-DSA) and EIP-8052 (Falcon-512) propose precompiled contracts to reduce PQC verification from ~1.8M gas to ~3,000 gas — a prerequisite for economically viable PQC transactions.
5.2 Consensus Layer: leanXMSS and leanV
Ethereum’s 500,000+ validators currently use BLS signatures, whose native mathematical aggregation compresses thousands of signatures into a single compact object. Post-quantum schemes lack this property. Without aggregation, un-aggregated PQC validator signatures would increase network bandwidth requirements to megabytes per slot, pricing out home validators.
- leanXMSS: Ethereum will replace BLS with leanXMSS, an optimised hash-based signature scheme for validators. Hash-based security relies on the security of hash functions rather than mathematical problems quantum computers are expected to solve.
- leanVM: A minimal zero-knowledge virtual machine (zkVM) generating SNARK-based proofs of ML-DSA and leanXMSS batch verifications, achieving approximately 250× compression of PQC validator signatures into a single compact proof. Preserves home-validator decentralisation.
- A leanSig Rust implementation and executable Python specification (leanSpec) are currently being used by approximately ten client teams building post-quantum consensus clients on dedicated devnets (pq-devnet-3, pq-devnet-4).
5.3 Emergency Hard Fork Contingency
In the event of an unexpected Q-Day event ahead of the migration schedule, Ethereum maintains a documented emergency contingency plan:
- Revert the blockchain to a pre-attack block height.
- Suspend all traditional EOA transactions.
- Force all unmigrated accounts to transition to smart contract wallets using hash-based signature verification.
6. Comparative Roadmap and Broader Ecosystem
6.1 Bitcoin vs. Ethereum: Side-by-Side
| Dimension | Bitcoin | Ethereum |
| Approach | BIP-360 (P2MR) + BIP-361 migration sunset | Strawmap: 4 cryptographic layers, ~7 hard forks |
| Key mechanism | Remove keypath spend; commit to Merkle root | Account abstraction (EIP-8141) + leanXMSS + leanVM |
| Governance | Decentralised, slow (SegWit took 8.5 yrs) | Coordinated via EF; faster but more centralised |
| Timeline | No formal deadline; BIP-360 testnet only | Core infrastructure targeted by 2029 (Hegotá fork H2 2026) |
| L2 coverage | Lightning (preserved in P2MR) | L2s NOT yet covered; separate effort needed |
| Status (May 2026) | BTQ testnet v0.3.0 live; mainnet TBD | pq-devnet-3/4 active; 10+ client teams |
6.2 Other Blockchains: Who Is Already Ahead
Bitcoin and Ethereum are not alone in facing the quantum transition — nor are they the fastest movers:
- QRL (Quantum Resistant Ledger): Has used hash-based signatures (XMSS) since its genesis block in 2018. Never used elliptic curve cryptography. The reference implementation for a fully quantum-resistant Layer 1.
- Algorand: Successfully implemented Falcon-512 in production in its September 2024 (v4.3.0) consensus upgrade, adding falcon_verify as a native AVM opcode. The strongest proof point that PQC signature verification can operate efficiently at scale on a production Layer 1.
- QANplatform and Hedera: Running NIST-approved Dilithium signatures in live environments.
- Zcash: Committed to a concrete 2027 timeline for quantum-proof wallets.
- XRPL: Assessed in May 2026 as potentially the furthest along of any top-10 blockchain in concrete post-quantum implementation, with native credential and identity features designed for compliance and institutional adoption.
The existence of these production deployments is significant: it demonstrates that PQC blockchain operation is not theoretical. The constraint for Bitcoin and Ethereum is not technical feasibility — it is governance coordination and migration scale.
7. Regulatory Context
Post-quantum migration is no longer exclusively a research concern, it has entered government policy frameworks:
- United States: US federal agencies faced an April 2026 deadline to submit post-quantum cryptography transition plans under NSM-10. Google has set its own internal 2029 migration deadline for all authentication infrastructure.
- European Union: The EU has set a target for critical infrastructure quantum-resistance by 2030. MiCA enforcement (July 2026) is creating a more stable regulatory environment for institutional crypto participation, which will intersect with PQC compliance requirements.
- BRICS: All 11 BRICS members are exploring alternative payments infrastructure, with cross-border wholesale CBDC projects doubling since 2022. PQC readiness is increasingly a consideration for sovereign digital currency design.
8. Key Risks and Open Questions
Despite significant progress, several fundamental challenges remain unresolved:
- The Satoshi coin problem: Approximately 1.1M BTC in Satoshi-era P2PK addresses cannot be voluntarily migrated. Any protocol-level decision about these coins (freezing, burning, or preserving) will be one of the most contested governance decisions in Bitcoin’s history.
- Bitcoin’s upgrade velocity: BIP-360 is a draft. Bitcoin Core has made no implementation progress toward mainnet. Given historical upgrade timelines of 7–8+ years for SegWit and Taproot, the window for orderly migration before Q-Day may be tight even if hardware timelines slip.
- Ethereum L2 coverage: The Strawmap does not currently address L2 wallet security. Most Ethereum users transact on L2s, meaning the base-layer PQC migration may secure settlement without securing user funds.
- Economic viability of PQC transactions: Even with precompiles, PQC transactions will be larger and more expensive than ECDSA transactions. User incentives to migrate voluntarily are unclear without regulatory pressure or visible quantum threat.
- ZK-proof system vulnerability: Many popular ZK-proof systems (SNARKs using elliptic curve pairings) rely on quantum-vulnerable mathematical assumptions. The migration of rollup proof systems — not just signatures — is a separate, complex engineering challenge.
- Supply chain attacks during migration: The migration period itself introduces risk. Hybrid systems (supporting both ECDSA and PQC) expand the attack surface and must be carefully designed to avoid downgrade attacks.
9. Conclusion
The quantum threat to blockchain cryptography is real, but the timeline remains measured in years rather than months. The Google whitepaper should be read as a precise calibration of the threat — not a declaration of imminent attack. The gap between demonstrated quantum hardware (approximately 28 logical qubits) and the attack threshold (1,200–1,450 logical qubits) remains enormous and involves challenging engineering problems.
What has fundamentally changed is the planning calculus. The 20× compression in resource estimates converts Q-Day from a generational concern into a medium-term institutional risk. Organisations holding significant Bitcoin or Ethereum (such as exchanges, custodians, sovereign wealth funds, tokenized asset platforms) should now treat PQC migration readiness as part of their risk management frameworks.
Both Bitcoin and Ethereum have technically credible migration paths. Bitcoin’s path (BIP-360 + BIP-361) is architecturally elegant but faces severe governance and adoption obstacles. Ethereum’s Strawmap is more structured and centrally coordinated, targeting 2029 for core infrastructure completion, but leaves the L2 ecosystem, where most value transacts, without a clear PQC plan.
The broader lesson from blockchains that are already quantum-resistant (Algorand’s Falcon-512 in production, QRL’s XMSS from genesis) is that the technology works at scale. The remaining challenges are human, not cryptographic: governance consensus, user migration incentives, and the politically charged question of what to do with the legacy coins of those who cannot or will not migrate.