In this article
- TL;DR
- BIP-360 Lays the Cryptographic Foundation
- BIP-361 Is the Proposal That Divided the Community
- The phased plan
- The Community Split
- Project Eleven Measured the Threat
- What a Post-Quantum Signature Would Cost
- What Institutional Holders Should Prepare For
- What Bitcoin Holders Should Do Now
- References
- What's Next
I work in the crypto self-custody space, and holders rarely ask me anymore whether quantum computers are real. They ask how much time they have and what happens to coins that are not moved in time. Two draft proposals give an answer to both, and neither answer is comfortable.
A March 2026 paper by Google Quantum AI researchers and others counts about 6.9 million BTC in outputs whose public key is already exposed (arXiv:2603.28846, figure 5). BIP-361 puts the same group at over 34% of all bitcoin as of 1 March 2026 (BIP 361). The paper credits accounts believed to be Satoshi Nakamoto's with about 1 million of them, which is about CHF 69 billion at the CHF 69,340 per bitcoin that mempool.space showed on 11 October 2026 (mempool.space).
BIP-360 and BIP-361 both sit in Bitcoin's BIP repository, which states that publication "does not indicate that it is a good idea, has community consensus, or that it is about to be adopted" (README). If you read our earlier piece on what Google's quantum paper means for holders, you know the cost estimates for an attack have fallen. These two BIPs are a concrete response, and BIP-361 has split the community.
TL;DR
BIP-360 adds a bc1z output type, Pay-to-Merkle-Root (P2MR), that hides the public key until you spend and so resists long-exposure quantum attacks. It adds no post-quantum signature. BIP-361 layers a deadline on top: after about 3 years, no new coins may go to legacy addresses; after 5 years, old-style signatures only work through a rescue protocol that proves you hold the wallet's seed. No such rescue is known for P2PK outputs, which hold Satoshi's coins. BIP-361 has no activation date. Taproot (bc1p) outputs expose the key today; P2WPKH (bc1q) outputs without address reuse do not.
BIP-360 Lays the Cryptographic Foundation
BIP-360 got its number on 18 December 2024 as Hunter Beast's "P2QRH" proposal for SegWit version 3 (BIP 360 as assigned). Ethan Heilman and Isabel Foxen Duke joined later. The authors renamed it Pay-to-Merkle-Root on 10 February 2026, and the BIP editors merged it a day later (pull request #1670).
The current text defines P2MR as a SegWit version 2 output with addresses that start with bc1z, encoded in bech32m (BIP 360). You have seen bc1q for SegWit and bc1p for Taproot. This is the next step in the same series.
What P2MR changes. Taproot (BIP-341) lets you spend two ways: a key path spend with a single signature, or a script path spend through a Merkle tree of scripts. The Taproot output itself contains a public key, so the key sits on-chain from the moment you receive. P2MR removes the key path and commits only to the Merkle root of the script tree. A spend still reveals the leaf script, which may contain a key, so the key is exposed while the spend waits in the mempool. BIP-360 calls that the short-exposure window and says it does not protect against attacks within it.
There is no signature scheme in it. The authors write that the proposal "does not include the introduction of post-quantum signature schemes" and name ML-DSA and SLH-DSA as examples to scrutinize (BIP 360). Both are NIST standards, FIPS 204 and FIPS 205, published on 13 August 2024 (NIST). ML-DSA is lattice-based. SLH-DSA is hash-based and rests on a different security assumption.
Earlier drafts did pick schemes. The authors dropped SQIsign, an isogeny-based scheme, on 20 January 2025, after the draft put its performance at "roughly 100,000 times slower than ECC" (BIP 360 commit). In my view that was the right call: a scheme that slows block validation that much is not usable on a full node.
The size cost of P2MR itself. A P2MR spend of a depth-1 tree needs a 135-byte witness, against 66 bytes for a Taproot key path spend (BIP 360). That is the price of hiding the key, before any post-quantum signature enters the picture.
For a deeper look at how Bitcoin's scripting system works and how Taproot laid the groundwork for this, see Technical Deep Dive.
BIP-361 Is the Proposal That Divided the Community
BIP-361 is the harder conversation. It asks what to do about coins that are not moved.
Jameson Lopp, Christian Papathanasiou, Ian Smith, Joe Ross, Steve Vaile and Pierre-Luc Dallaire-Demers wrote BIP-361, which was assigned on 11 February 2026 and merged on 14 April 2026 (BIP 361, pull request #1895). Its motivation section says over 34% of all bitcoin had a revealed public key on 1 March 2026.
The Google Quantum AI paper breaks that pool down. P2PK outputs, the format of the earliest mining rewards, hold over 1.7 million BTC (arXiv:2603.28846). These outputs store the public key directly on-chain. The paper says accounts believed to be Satoshi's hold about 1 million of those. The rest of the roughly 6.9 million sits in Taproot outputs, which show the key by design, and in older hashed types whose key was revealed by reuse (arXiv:2603.28846, figure 5).
The phased plan
Phase A starts 160,000 blocks, about 3 years, after BIP-361 activates (BIP 361). From then on, legacy scripts may only send to post-quantum scripts. You can still spend from a legacy address, but nobody can send new coins into one. That pushes every new deposit toward the post-quantum output types.
Phase B starts 2 years after Phase A, so about 5 years after activation (BIP 361). Nodes then tighten the rules for ECDSA and Schnorr spends: a spend must also pass a quantum-safe rescue protocol that a quantum attacker could not satisfy.
The text changed three days after the merge. The version merged on 14 April 2026 said Phase B "renders ECDSA/Schnorr spends invalid" and left recovery to a later Phase C, a zero-knowledge proof of a BIP-39 seed phrase (BIP 361 as merged). A commit on 17 April 2026 removed Phase C and folded the rescue into Phase B (BIP 361 commit). The example it now gives is a proof of the BIP-32 parent key behind a wallet that used hardened derivation, something a quantum attacker would be very unlikely to know.
That rescue does not reach P2PK outputs. BIP-361 says "it's not currently believed possible to construct a rescue protocol for P2PK UTXOs", and its authors support a separate proposal called Hourglass for them (BIP 361). Satoshi's coins fall in that bucket.
On deployment timing. A pre-merge draft still hosted at bip361.org listed a start time of epoch 1798761600, midnight on 1 January 2027 UTC, for mainnet, testnet4 and signet (bip361.org). Lopp removed that section before the merge, writing that he had "added the deployment section as a placeholder" and that it "doesn't really add value" (pull request #1895). The text specifies the phases but no activation parameters. BIP-361 also lists its requirement as "TBD Post Quantum Signature BIP", so it cannot start before a signature proposal exists.
For context on how dormant coins, inheritance planning and lost access interact with this proposal, see Bitcoin Inheritance.
The Community Split
The opposition to BIP-361 is on record. CoinDesk quoted an X user who called it "highly authoritarian and confiscatory" and wrote that "Upgrade should be 100% voluntary" (CoinDesk, X post). The argument is that restricting coins that have not moved, whatever the reason, breaks the rule that whoever holds the keys controls the coins.
The authors argue the opposite. Their text calls the plan "defensive" and puts it this way: "Quantum recovered coins only make everyone else's coins worth less. Think of it as a theft from everyone." (BIP 361). Coins in P2PK outputs stay safe only as long as no machine can derive their private keys. Leaving them exposed hands the question of who controls them to whoever builds that machine first.
In my view neither side is obviously wrong. The freeze step needs a coordinated decision about other people's coins, and Bitcoin has avoided that so far. Whether the threat justifies the precedent is a question the community is still working through.
Project Eleven Measured the Threat
The debate got a data point on 24 April 2026. Project Eleven awarded its 1 BTC Q-Day Prize to researcher Giancarlo Lelli, who broke a 15-bit elliptic curve key on publicly accessible quantum hardware (Project Eleven on X).
Project Eleven called it a "512x jump from the previous public demonstration" (Project Eleven on X), which was a 6-bit break in September 2025 (The Quantum Insider). A 15-bit key is far smaller than the 256-bit keys Bitcoin uses, and CoinDesk noted the feat is "far from threatening bitcoin's 256-bit elliptic curve security" (CoinDesk). In my view the result matters as a trend, not as an attack: the same kind of curve, on real hardware, at a key size that grew from 6 to 15 bits in seven months.
For a full breakdown of what this class of quantum research means for Bitcoin holders specifically, see our earlier piece on quantum computing and Bitcoin.
What a Post-Quantum Signature Would Cost
In my view, size is why a signature proposal is harder to write than BIP-360.
An ML-DSA-44 signature is 2,420 bytes (NIST FIPS 204, table 2). A Schnorr signature is 64 bytes and a DER-encoded ECDSA signature is up to 72 bytes (BIP 340). The post-quantum signature is more than 33 times larger than the biggest ECDSA one.
Under SegWit, a witness byte counts as 1 weight unit and a non-witness byte as 4, and a block may hold 4,000,000 weight units (BIP 141). If a future soft fork put ML-DSA-44 signatures in the witness, each would weigh 2,420 units, or 605 virtual bytes.
Here is the arithmetic in francs. On 11 October 2026 mempool.space suggested 4 sat/vB for the next block and showed CHF 69,340 per bitcoin (fees, price).
| Signature | Size | Virtual bytes | Fee at 4 sat/vB | In CHF on 11 Oct 2026 |
|---|---|---|---|---|
| Schnorr (BIP 340) | 64 bytes | 16 | 64 sat | about CHF 0.04 |
| ML-DSA-44 (FIPS 204) | 2,420 bytes | 605 | 2,420 sat | about CHF 1.68 |
That is for the signature alone, on a quiet day. Fees scale with demand for block space, so the same signature costs more when blocks are full. Which scheme Bitcoin adopts, and how, is for a proposal that does not exist yet.
What Institutional Holders Should Prepare For
You need to do nothing today. Both proposals are drafts, BIP-361 has no activation parameters, and any soft fork goes through review and testing before it goes live.
If either proposal moves ahead, custodians and institutional holders will need to plan key rotation and address migration. Every UTXO you want protected must move to a new output type. For an institution holding Bitcoin across thousands of addresses, that is a large operational job.
Swiss banks and custodians already have a date recommended by their regulator. FINMA's guidance 05/2026 of 9 July 2026 recommends that supervised institutions draw up a post-quantum cryptography roadmap by mid-2027 at the latest, and it cites NIST FIPS 203, 204 and 205 as the quantum-safe algorithms it means (FINMA). The guidance names distributed ledgers among the technologies an institution's inventory should cover. ML-DSA, the first example BIP-360 names, is FIPS 204.
In my view the institutional question has already shifted from "if" to "when". A Swiss custodian that writes its FINMA roadmap in 2027 will have to say what it does with Bitcoin keys.
For practical guidance on securing your Bitcoin today, regardless of quantum timelines, see the self-custody guide and how seed phrases resist brute force.
What Bitcoin Holders Should Do Now
The quantum threat is real but not imminent. Four habits help now.
Check your address types. BIP-360 lists P2PK and Taproot outputs as vulnerable to long-exposure attacks, because the public key sits on-chain (BIP 360). Addresses starting with bc1p are Taproot. Addresses starting with bc1q (P2WPKH), 1 (P2PKH) or 3 (P2SH) hide the key behind a hash until you spend or reuse them. P2PK coins, mostly from the 2009 to 2010 mining era, carry the highest exposure (arXiv:2603.28846). See Wallets: Staying Secure for a full breakdown of address types.
Avoid address reuse. Spending from an address puts its public key on-chain, and coins left at that address are exposed from then on (BIP 360). A fresh address for each receive limits that exposure.
Keep your seed phrase secure. If BIP-361 activates and you miss both deadlines, the rescue protocol it describes relies on proving knowledge of the BIP-32 parent key that your seed produces (BIP 361). Lose the seed and those coins stay locked. That holds whatever happens with quantum computing.
Stay informed, not alarmed. BIP-361 cites McKinsey for an estimate that a cryptographically relevant quantum computer could exist as early as 2027 to 2030 (BIP 361). Google set 2029 as the timeline for its own post-quantum migration (Google). Migration needs every holder to act, so a few years is less time than it sounds.
For more on address privacy and reducing on-chain exposure, see Bitcoin Privacy.
References
- BIP-360: Hunter Beast, Ethan Heilman, Isabel Foxen Duke. "Pay-to-Merkle-Root (P2MR)." Draft, version 0.12.1. Assigned 18 December 2024. github.com/bitcoin/bips, project site bip360.org
- BIP-361: Jameson Lopp, Christian Papathanasiou, Ian Smith, Joe Ross, Steve Vaile, Pierre-Luc Dallaire-Demers. "Post Quantum Migration and Legacy Signature Sunset." Draft. Assigned 11 February 2026. github.com/bitcoin/bips
- arXiv:2603.28846: Babbush et al. "Securing Elliptic Curve Cryptocurrencies against Quantum Vulnerabilities: Resource Estimates and Mitigations." March 2026. arxiv.org
- NIST Post-Quantum Cryptography Standards: FIPS 203 (ML-KEM), FIPS 204 (ML-DSA), FIPS 205 (SLH-DSA). Published 13 August 2024. csrc.nist.gov
- FINMA-Aufsichtsmitteilung 05/2026: Quantum Computing. 9 July 2026. finma.ch
- Project Eleven Q-Day Prize: Giancarlo Lelli, 1 BTC for a 15-bit ECC key on quantum hardware. 24 April 2026. The Quantum Insider
- CoinDesk, 15 April 2026: "Bitcoin developers are trying to build quantum defenses. Your coins could pay the price." coindesk.com
- CoinDesk, 24 April 2026: "Researcher wins 1 bitcoin bounty for 'largest quantum attack' on underlying tech." coindesk.com
What's Next
This article covered what Bitcoin developers are building. For the underlying threat, how Google's paper changed the math, and what it means for your specific wallet, read Quantum Computing and Bitcoin: What Holders Must Know.
To understand the technical foundations that make these proposals possible, start with Technical Deep Dive. For practical steps on securing your Bitcoin today, see Wallets: Staying Secure.
Frequently Asked Questions
What is BIP-360?
bc1z. It protects against long-exposure quantum attacks only. It adds no post-quantum signature scheme; the authors name ML-DSA and SLH-DSA as examples and leave signatures to a later proposal. BIP-360 got its number on 18 December 2024, and its current authors are Hunter Beast, Ethan Heilman and Isabel Foxen Duke.What is BIP-361?
Will Satoshi's Bitcoin be frozen?
How large is a post-quantum signature compared with today's?
When will Bitcoin become quantum-resistant?
What does this mean for me as a Bitcoin holder?
bc1p) outputs show the public key on-chain, so they are exposed to long-exposure attacks today (BIP 360). P2WPKH (bc1q) and other hashed types keep the key hidden until you spend or reuse the address. Keep your seed phrase safe: the BIP-361 rescue idea relies on proving knowledge of the BIP-32 parent key behind your wallet.