No quantum computer can break Bitcoin’s signatures today, but the risk is already shifting toward dormant coins, custodians and bridges. Between 6 and 6.9 million BTC are tied to exposed public keys, and developers are now debating migration rules rather than the threat itself. This analysis separates the numbers from the speculation.
๐ Key takeaways
- No quantum attack is underway; the risk centers on migrating coins and infrastructure.
- 6 to 6.9 million BTC are associated with public keys already visible on-chain, depending on the estimate.
- Google and Caltech sharply reduce the hardware needed, though their architectures differ.
- The BIP 360 and BIP 361 proposals remain under discussion; neither has been adopted.
A threat that depends on exposed public keys
Quantum risk does not target Bitcoin transactions in general today. The network relies on ECDSA and, since 2021, on Schnorr, both built on the elliptic curve discrete logarithm problem. Current supercomputers would take trillions of years to reverse it, while a cryptographically relevant quantum computer (CRQC) could do so in hours or days. Bitcoin Suisse stresses that a quantum computer is not simply a faster classical machine: qubits, entanglement and interference enable new kinds of operations, and Shor’s algorithm is the direct threat to signatures.
Exposure depends above all on whether public keys are visible. A revealed public key would allow a future adversary with sufficient hardware to recover the private key and forge a spend. P2PK outputs from the Satoshi era keep the public key directly on-chain. Address reuse reveals it after the first outgoing payment, and the P2TR (Taproot) format, adopted in 2021, exposes it by default. The Human Rights Foundation (HRF) notes that anyone who has received bitcoin on a Taproot address is affected. These cases fall under so-called long-range attacks, which could target coins already held at rest.

6 million BTC, and estimates that diverge
Exposure estimates do not line up. CoinDesk, drawing on Glassnode data, reports that more than 6 million BTC, or 31.2% of circulating supply, was linked to public keys already visible in October. QuSecure, commenting on Google’s work, puts the figure at 6.9 million BTC. The table below gathers the main estimates.
| Source | Exposed volume | Share of supply |
|---|---|---|
| CoinDesk (Glassnode) | More than 6 million BTC | 31.2% |
| QuSecure | 6.9 million BTC | n.a. |
| Chaincode (May 2025) | About 6.26 million BTC (range of 4 to 10 million) | 20% to 50% (range) |
| HRF | 6.51 million BTC | n.a. |
| Citi Institute (January 2026) | n.a. | About 25% |
HRF uses 6.51 million BTC, worth more than $718 billion, and splits it into 4.49 million BTC ($495 billion) that holders could move, and 1.72 million BTC ($188 billion) considered dormant or inaccessible. However, 4.49 + 1.72 = 6.21 million, which leaves 300,000 BTC unexplained against the stated total. The sources share no common methodology that reconciles these figures, which may also reflect different dates and valuations. CoinDesk adds that these volumes cannot tell us how many bitcoins are lost, dormant or impossible to move.
Google and Caltech cut the hardware bill
Google Quantum AI published a paper on March 31, co-written with Justin Drake of the Ethereum Foundation and Dan Boneh of Stanford. According to QuSecure, on a fast-clock superconducting architecture, deriving a private key from a public key would take 9 to 12 minutes. This so-called “on-spend attack” would target a transaction still waiting for confirmation. The same document estimates 18 to 23 minutes for coins whose keys are already visible, where the attacker has no race against confirmation to win.
- Hardware requirement: fewer than 500,000 physical qubits, about 1,200 logical qubits and 90 million Toffoli gates.
- Reduction: roughly 20 times lower than earlier estimates, which were expressed in millions of physical qubits.
A second paper, released the same day by Caltech and Oratomic, proposes an architecture built on 10,000 reconfigurable atomic qubits and qLDPC error-correcting codes. Those codes would require 161 times fewer physical qubits than surface codes for equivalent error correction. Existing atomic arrays have reportedly demonstrated up to 6,100 atoms, shrinking the gap between demonstrated hardware and a cryptographically relevant machine from three orders of magnitude to a factor of roughly four. The two approaches rest on different architectures, so their estimates are not directly comparable. Google validated its results with zero-knowledge proofs and did not publish executable attack details, given their dangerous nature.
Timelines and the migration debate
The timeline remains uncertain. The Chaincode report notes that one third of respondents to a global expert survey put the odds at 50% or more that a CRQC able to break Bitcoin’s cryptography appears before 2030-2035. It also recalls that national and governmental guidance foresees deprecating elliptic curves by 2030 and banning them by 2035. HRF, in its October 31, 2025 report, says some experts point to arrival within the next five years.
“Sufficiently powerful quantum computers (and other ECDLP breaks) are hypothetical, and if they happen, there will likely be a long series of incremental breakthroughs that give us time for more fundamental solutions.”
Pieter Wuille, Bitcoin Core contributor
Positions diverge sharply. Sho Suigara, CEO of Blocq, considers five to ten years a reasonable horizon while advising against panic. Terry Rudolph, CEO of PsiQuantum, warns that a phase transition could arrive as abruptly as the AI boom. At the Presidio Bitcoin Quantum Summit, the share of participants unsure about a significant quantum threat fell from 25% to 8% after two days of discussion, while the share expecting a CRQC within five to twenty years rose from 49% to 69%.
Developers are also debating what happens to coins that never migrate. In April, Jameson Lopp and five co-authors proposed BIP 361, which would gradually phase out vulnerable signature methods and eventually restrict transactions from non-upgraded wallets. Lopp argues that freezing dormant coins may be preferable to letting future quantum attackers steal them. Adam Back, CEO of Blockstream, favors voluntary migration and warns against an imposed migration through a predetermined freeze. BIP 360, which proposes a new output type to reduce exposure to certain attacks, also remains a draft. Neither approach has been adopted.
“The clock that matters is migration.”
Leo Fan, CEO of Cysic
Fan adds that coins that never move are a social problem more than a technical one. On closing the old path, he is equally blunt: if it stays open forever, attackers will use it.
Custodians, NIST standards and infrastructure
The transition goes well beyond individual wallets. Custodians often rely on multi-party approval rather than a single private key, and they will need post-quantum signatures. Fan notes that post-quantum signatures for a single signer are standardized, but threshold and multisig versions remain at the research stage. The NIST launched a call for proposals on multiparty threshold cryptography in January. In May, BitGo and Silence Laboratories ran a simulated post-quantum transaction using secure multiparty computation, with no production rollout announced.
- NEAR introduced post-quantum transaction signing in July.
- Algorand introduced native Falcon-1024 accounts in August.
- Rootstock, a Bitcoin sidechain, plans quantum-resistant signatures in 2028, along with changes to PowPeg, its hardware security modules and protections on the Bitcoin side. These remain proposals, and Rootstock cannot upgrade the underlying Bitcoin protocol on its own.
In August 2024 the NIST finalized three standards: FIPS 203 ML-KEM for key encapsulation, FIPS 204 ML-DSA for signatures and FIPS 205 SLH-DSA for hash-based signatures. ML-KEM covers encrypted channels, storage and wallet-to-wallet communication, while ML-DSA and SLH-DSA are the most direct standardized comparators for transaction signatures. A hybrid migration requires both a classical and a post-quantum verification for every authorization. ML-DSA-44, ML-DSA-65 and ML-DSA-87 signatures weigh roughly 2.4 KB, 3.3 KB and 4.6 KB, before script and encoding overhead. HRF warns that this size increase could substantially grow the chain, which already faces scaling difficulties.
Other components are exposed too. Exchanges, custodians, node operators and wallet applications still depend on TLS, API authentication and classical internal communications. Many wallet software products, accounting tools and portfolio trackers also store users’ public keys, so a breach could, in a world where Shor’s algorithm works, facilitate theft of the corresponding funds. Mining looks less exposed: Grover’s effect halves the theoretical security margin of hash functions, but the potential quadratic advantage of quantum miners would not yet be enough to outpace highly optimized ASICs. Chaincode nonetheless flags double-spend risks from small and mid-sized miners, and centralization risk if quantum mining became dominant.
Technical options remain exploratory: Lamport signatures, quantum-secure Taproot scripts, payments to a quantum-resistant hash and a commit-delay-reveal mechanism. Chaincode proposes two parallel timelines: about two years for contingency measures that can be deployed quickly, and about seven years for a full transition.
Conclusion: anticipate rather than react
The quantum threat to Bitcoin is not yet measured in attacks but in readiness. Scenarios diverge: if CRQCs arrive late, voluntary migration and NIST standards may suffice; if they appear within five to ten years, the fate of dormant coins and centralized infrastructure becomes the priority. No migration standard is currently built into the protocol, and no single actor can impose adoption on every user, custodian, bridge and infrastructure operator.
The sources report no price or volume movement attributable to this research or to this exposure. For now, the market has not priced in a measurable quantum risk, which says nothing about its relevance for long-term holders.
Sources
- CoinDesk, November 22, 2026
- QuSecure, analysis of the Google paper
- Chaincode Labs, Bitcoin post-quantum report
- Bitcoin Suisse, The quantum threat to Bitcoin
- Human Rights Foundation, report of October 31, 2025
- PubMed Central, article PMC13556152
Cet article est publiรฉ ร titre informatif et รฉducatif. Il ne constitue en aucun cas un conseil en investissement. Faites vos propres recherches (DYOR) avant toute dรฉcision.


