Taproot in 2026: Fork Fights, Real Uses, and the Quantum Clock

Over the weekend of August 8 and 9, Bitcoin did something it had not done in years: it repelled a soft fork in real time, and the whole thing was over almost before most holders noticed. A proposal called BIP-110, designed to temporarily choke off the arbitrary data that now rides inside Bitcoin transactions, opened its miner-signaling window with just 2.53% support. A handful of miners tried to enforce it anyway, split onto their own chain, mined exactly two blocks in about eight hours, and then stalled. Roughly 99.85% of the network’s hashpower never left the main chain, according to CoinDesk.

That episode is a useful place to start a story about Taproot, because the fight was really about Taproot. BIP-110 targeted the flexible witness data and scripting that Bitcoin’s 2021 Taproot upgrade made cheap and expressive. Nearly five years after activation, Taproot is the plumbing under most of what is interesting on Bitcoin right now: cheaper multisig, dollar stablecoins moving over Lightning, a new privacy standard, and, less comfortably, tens of millions of Ordinals inscriptions that a large camp of Bitcoiners would rather never existed. It is also, according to a piece of Google research from March 2026, the address type most exposed to a future quantum computer.

This is a status report on where Taproot actually stands in 2026: what it changed, where it is used, why it keeps starting fights, and why its greatest strength (revealing less while doing more with a single flexible output) has become the center of Bitcoin’s most serious long-term security debate. BTC trades near $64,000 as of this writing, about 49% below its October 2025 record of $126,080, per CoinGecko, and the protocol questions swirling around Taproot are, increasingly, part of that price conversation too.

What Taproot Actually Changed

Taproot shipped as three linked Bitcoin Improvement Proposals, and it helps to keep them straight. BIP 340 introduced Schnorr signatures, authored by Pieter Wuille, Jonas Nick and Tim Ruffing. BIP 341 defined Taproot itself, the Pay-to-Taproot (P2TR) output type and the Merkelized Alternative Script Tree, or MAST. BIP 342 defined Tapscript, the scripting rules for spending those outputs, as documented in the bitcoin/bips repository.

The core idea is elegant. Before Taproot, Bitcoin used ECDSA signatures, and complex spending conditions (a 2-of-3 multisig, a timelock, a Lightning channel) had to be spelled out on-chain when the coins moved, exposing the whole script for everyone to see. Schnorr signatures changed the math. They are linear, which means several public keys can be added together into one aggregate key and several signatures into one aggregate signature. A group of signers can produce a single Schnorr signature that looks identical to a signature from one person.

MAST does the same trick for scripts. Instead of revealing every possible spending condition, a Taproot output commits to a Merkle tree of them and reveals only the branch that gets used. Combine the two and you get a P2TR output that, in the common case, looks like a plain single-signature payment even when a sprawling contract sits behind it. There are two ways to spend it: the key path (a single Schnorr signature against the aggregated key, cheap and private) and the script path (revealing one MAST branch when the contract logic is actually needed).

The visible signal is the address. Taproot addresses start with bc1p, distinct from the bc1q of native SegWit and the leading 1 of legacy addresses. The promised payoffs were three: smaller and cheaper complex transactions, better privacy (a multisig vault is indistinguishable from a normal wallet on the key path), and more room for builders. Nearly five years on, all three have partly arrived, in ways nobody fully predicted.

Activation, Signaling, and the Rules of a Soft Fork

Taproot’s activation is worth remembering in 2026 because the mechanism is back in the news. After the bruising 2017 SegWit fight, developers wanted a calmer path, and they got one with a process called Speedy Trial. Miners signaled readiness inside a fixed window; the upgrade locked in at block 687,284 in June 2021 after clearing more than 90% miner support, and it activated at block 709,632 on November 14, 2021, as CoinDesk reported at the time. By Bitcoin standards it was drama-free: no chain split, no competing client war, and the change was backward compatible.

The reason that matters is the logic of a soft fork. A soft fork tightens the rules, so old nodes still accept the new blocks, but it only works if an overwhelming majority of miners and economic actors go along. If a large minority refuses, you get two chains. Taproot had near-unanimous support, so there was nothing to split. That is precisely the test any future change has to pass, and it is the test BIP-110 walked straight into and failed.

The BIP-110 Fork That Just Fizzled

BIP-110, the Reduced Data Temporary Softfork proposed by developer Dathon Ohm, was an attempt to use that same signaling machinery to walk Taproot back a little, as laid out on bips.dev. For one year it would have capped output scripts at 34 bytes (83 for OP_RETURN), limited push and witness items to 256 bytes, invalidated the Taproot annex, shrunk the Taproot control block, and banned several Tapscript opcodes. The target was the arbitrary data (mostly Ordinals inscriptions and token metadata) that Taproot’s roomy witness made cheap to store. Existing coins would have been grandfathered in.

It never had the votes. Its modified activation used a low 55% miner threshold, which critics warned was itself dangerous, and even that was a fantasy: signaling sat near 2% for weeks. When the mandatory window opened around August 8 at block 961,632, only 51 of the prior period’s blocks (2.53%) had signaled. A minority of miners, led by Ocean’s pool through its DATUM protocol, enforced the new rules anyway and rejected the first non-signaling block, which had been mined by AntPool, splitting onto a separate chain. That chain produced two blocks in roughly eight hours while the main chain rolled on for 48, then it froze. Holding barely 0.15% of hashpower against Bitcoin’s full difficulty, the breakaway chain would need close to a year to reach its first difficulty adjustment, and by some estimates far longer, per CoinDesk.

The reaction from Bitcoin’s loudest voices was less about data policy than about governance. Michael Saylor, executive chairman of Strategy and steward of the largest corporate BTC treasury, posted on X that “Bitcoin worked exactly as designed. BIP-110 was free to fork, and the network was free not to follow,” adding that “consensus is earned, not declared,” in remarks compiled by The Crypto Times. He was not alone in opposing it. Blockstream chief executive Adam Back had already called the design “not the way,” arguing it “breaks multiple things, doesn’t have tech nor ecosystem consensus,” while Casa co-founder Jameson Lopp labeled the 55% threshold “reckless” for how much it raised the odds of exactly the split that followed, both via Bitcoin.com News.

The takeaway for Taproot is that the market has, for now, decisively refused to restrict what it enables, even as the underlying disagreement about what Bitcoin blockspace is for remains unsettled. And a separate, messier fork looms. LayerTwo Labs founder Paul Sztorc has an opt-in Drivechain hard fork pencilled in for around August 21 that would airdrop a new eCash asset to holders and, controversially, reassign long-dormant coins linked to Satoshi, a plan Fidelity Digital Assets and others have opposed, as CoinDesk detailed. Unlike BIP-110, it is an opt-in new chain rather than a split of Bitcoin itself.

Taproot Adoption by the Numbers

For all the noise, Taproot’s on-chain footprint is moderate and, revealingly, cyclical. Its transaction share peaked above 40% in early 2024, when Ordinals inscriptions and the April 2024 launch of the Runes token protocol filled blocks with Taproot spends. As that speculative demand cooled, usage fell back to roughly a fifth of transactions through 2026, according to public adoption trackers. The rest of the network still runs mostly on native SegWit (bc1q), with legacy addresses shrinking every year.

Two things explain the slow organic climb. First, Taproot’s fee savings over native SegWit are real but modest, on the order of 15% to 16% for simple spends, far smaller than SegWit’s savings over legacy, so there is less pressure to switch. Second, support for bc1p is still patchy: some exchanges and services accept Taproot withdrawals but cannot yet generate bc1p deposit addresses, which slows the flywheel. As custody and wallet software catch up, the baseline should keep grinding higher even without another inscription mania. Organic Taproot demand, in other words, now depends less on speculation and more on the applications built on top of it.

Multisig Grows Up: MuSig2, FROST, and Threshold Signatures

The clearest place Taproot delivered on its promise is multisig, the shared-custody setups that protect exchanges, funds and, increasingly, ordinary savers. Before Taproot, a 2-of-3 vault published all three public keys and a script every time it spent, which was expensive and told the world exactly how the coins were guarded. Schnorr key aggregation changed that through MuSig2, standardized as BIP 327 by Jonas Nick, Tim Ruffing and Elliott Jin.

With MuSig2, the signers combine their keys off-chain into one aggregate key and produce one signature. On-chain, a MuSig2 vault spending on the key path is indistinguishable from a single-signature wallet. BitGo, which runs institutional custody at scale, published field numbers that make the efficiency concrete: a MuSig2 key-path input weighs about 57.5 virtual bytes against 104.5 for a native SegWit multisig, roughly a 45% reduction, and the design lets the last signer use a deterministic nonce so a hardware security module can sign without keeping state, as Bitcoin Optech documented. Casa co-founder Jameson Lopp has framed the privacy win bluntly: with Taproot multisig, wallets “no longer have to reveal what keys were participating” and can “blend in with the crowd.”

The frontier is threshold signatures, where any k-of-n subset can sign without an interactive setup that reveals the policy. FROST (Flexible Round-Optimized Schnorr Threshold signatures), from academic work by Chelsea Komlo and Ian Goldberg and standardized by the IETF as RFC 9591 in 2024, is the leading scheme; a Bitcoin-specific draft, BIP 445, was assigned in early 2026 but remains a work in progress. Lightspark’s Spark, a Bitcoin layer-two, already uses FROST for a 1-of-n operator model whose settlement transactions look like ordinary Taproot spends.

The parallel with Ethereum is instructive. Where Ethereum spent 2026 turning ordinary accounts into programmable smart wallets through set-code transactions under EIP-7702, Bitcoin got a lot of the same expressiveness (shared control, spending policies, privacy) from Taproot’s signature math, without touching its account model at all. Not every custodian has jumped: some collaborative-custody providers still prefer explicit multi-leaf Tapscript over interactive Schnorr signing, citing the operational risk of nonce reuse. But the direction of travel is clear.

Stablecoins on Bitcoin: Taproot Assets and RGB

The most surprising thing to grow out of Taproot is a stablecoin rail. Taproot Assets, built by Lightning Labs, uses Taproot outputs to anchor tokens issued through client-side validation, so the tokens ride on Bitcoin and move over the Lightning Network. In March 2026, Tether’s USDT went live on Lightning via Taproot Assets as USDT-L, a wrapped representation custodied by Cantor Fitzgerald. Channel opens that had cost more than $70 at the 2023 inscription peak were suddenly running around $0.12.

Lightning Labs chief executive Elizabeth Stark has long argued that Taproot Assets could usher in a “Bitcoin Renaissance” and cement Bitcoin as the “internet of money,” a claim she first made at the protocol’s 2023 mainnet debut, per The Block. The 2026 version of that thesis has a sharp regulatory edge. USDT was pulled from Europe’s licensed exchanges as the bloc’s crypto rulebook took full effect, because Tether never sought an e-money authorization, per crypto.news, so a Bitcoin-native rail that individuals can use through self-custody is more than a technical curiosity. A rival protocol, RGB, led by Maxim Orlovsky’s LNP/BP Standards Association, reached mainnet in 2025 and is racing to host stablecoins of its own using a fully peer-to-peer form of the same client-side-validation idea.

Privacy: Silent Payments and Why They Need Taproot

Another Taproot-dependent advance arrived without any consensus change at all. Silent Payments, specified as BIP 352 by Josie Baker, Ruben Somsen and Sebastian Falbesoner, let someone publish one static address and receive unlimited payments to it without any of those payments looking linked on-chain, as set out in the bitcoin/bips repository. The sender derives a unique one-time output for each payment using a shared secret computed from the transaction’s own inputs, so there is no reused address and no notification transaction cluttering the chain.

The catch, and the Taproot connection, is that the output must be a Taproot (P2TR) output to blend in. A silent payment lands as an ordinary-looking bc1p spend, indistinguishable from any other Taproot transaction, which is the entire point. Wallets including Sparrow, Cake and the BitBox02 hardware device added support across 2024 and 2025, and dedicated scanning servers emerged to handle the heavier work of detecting incoming payments. It is a reminder that Taproot’s biggest gift to Bitcoin may be optionality: a single flexible output type that later standards can build on without asking miners for permission.

Ordinals, Runes, and the Data Wars

Not every use of that optionality was welcome. When developer Casey Rodarmor launched the Ordinals protocol in January 2023, he used Taproot’s expanded witness space to inscribe arbitrary data (images, text, even small programs) directly onto individual satoshis. Runes, Rodarmor’s fungible-token protocol, followed on the April 2024 halving and stored token state natively in Bitcoin’s UTXO model rather than bloating it with text the way earlier BRC-20 tokens had. Together they generated tens of millions of inscriptions and, for stretches of 2024, more than half of all Bitcoin transactions.

For miners, the inscription boom was a windfall. Block space suddenly had bidders beyond simple payments, and fee revenue spiked at exactly the moments the network needed it, a dynamic we unpack in our guide to hashprice, mining’s master metric. For a large and vocal camp of Bitcoiners, it was vandalism: a monetary network turned into a costly, permanent image board. That disagreement produced the OP_RETURN wars. Bitcoin Core version 30, released in October 2025, raised the default limit on OP_RETURN data from 83 bytes toward 100,000 and allowed multiple such outputs, prompting Bitcoin Knots maintainer Luke Dashjr to call the change “malicious code” and driving a surge of node operators toward the rival Knots implementation, per CoinDesk.

BIP-110 was the next escalation of that same fight, and its failure this month settled little except the immediate question of whether a miner minority could force the issue. It could not. But the deeper argument, whether Bitcoin is a settlement network that should resist arbitrary data or a base layer that should host whatever pays the fee, is exactly the argument that will shape which future Taproot-enabled applications thrive.

The Covenant Renaissance and Its Stall

If Taproot opened the door to richer Bitcoin contracts, a cluster of 2026 proposals wants to open it wider through covenants, rules that constrain how coins can be spent in the future (think vaults that force a withdrawal delay, or congestion-control batching). The trouble is that activating any of them requires the same miner and economic consensus that Taproot earned and BIP-110 could not.

OP_CTV (BIP 119), championed by Jeremy Rubin as the safe, non-recursive covenant, has an activation client and a signaling window that opened in March 2026, but miner support has languished near 2% against a 90% threshold, leaving activation before its March 2027 timeout unlikely, per a BlockEden analysis. OP_CAT (BIP 347), which would re-enable script concatenation and, combined with Schnorr, unlock recursive covenants, has a complete specification and no activation path at all. LNHANCE bundles several Lightning-focused primitives with no roadmap yet.

The most consequential development routes around the stalemate entirely. BitVM2 needs no soft fork; it uses optimistic fraud proofs to verify computation off-chain and settle disputes on Bitcoin. It powers Citrea, which its team describes as the first production zero-knowledge rollup built directly on Bitcoin and which launched mainnet on January 27, 2026, as The Block reported. These systems inherit Bitcoin’s security only as strongly as their bridges do, which is why the design of Bitcoin layer-two bridges, covered in our explainer on how cross-chain trust actually works, matters as much as the cryptography.

Blockstream researcher Andrew Poelstra, a co-author on several Taproot BIPs, has pushed back on the idea that Bitcoin should simply ossify, arguing that the network is already evolving and that developers “need to argue passionately and correctly” for the changes worth making, in remarks reported by Bitcoin Magazine.

The Quantum Reckoning

The debate that could eventually overshadow all of this is quantum computing, and Taproot sits uncomfortably at its center. In March 2026, Google Quantum AI published research suggesting that breaking the elliptic-curve cryptography behind Bitcoin signatures might need fewer resources than previously feared: on the order of 1,200 to 1,450 logical qubits and under 500,000 physical qubits, down from earlier estimates in the millions, as covered by CoinDesk. Google’s own write-up framed the disclosure as deliberate and partial, meant to prompt migration without handing attackers a blueprint, and put its milestone for useful quantum systems around 2029, on the Google Research blog.

Here is why Taproot is the sharp end. Legacy and native SegWit addresses hide the public key behind a hash until the coins are first spent, so a quantum attacker sees only the hash while the coins sit still. A Taproot output, by contrast, reveals the full public key on-chain by default, from the moment it is created. That is what makes the key-path spend so efficient, and it is also what makes every Taproot output quantum-exposed from birth rather than only at spend time. By common estimates, roughly a third of all mined BTC (about 6.9 million coins, counting reused addresses and the earliest pay-to-public-key coins) already sits in an exposed state.

The named reactions split predictably. On-chain analyst Willy Woo urged holders to treat hash-protected addresses as an interim safe harbor but stayed calm on the endgame: “Quantum will not break BTC because BTC will adapt.” Ethereum Foundation researcher Justin Drake said the Google paper sharply raised his odds of a so-called q-day this decade: “My confidence in q-day by 2032 has shot up significantly.” Binance founder Changpeng Zhao waved off the panic: “All crypto has to do is upgrade to Quantum-Resistant Algorithms. So, no need to panic.” The Woo comments were reported by Yahoo Finance, and the Drake and Zhao remarks by CoinDesk. The disagreement is less about whether Bitcoin can adapt than about how much time it has to coordinate the change.

Bitcoin’s Post-Quantum Plan: BIP-360 and BIP-361

The technical response is already drafted. BIP-360, authored by Hunter Beast, Ethan Heilman and Isabel Foxen Duke, defines a new Pay to Merkle Root (P2MR) output type that removes the quantum-vulnerable key-path spend, forcing spends down a hashed script path and leaving room to add post-quantum signature schemes later; it was merged into the bitcoin/bips repository in February 2026 as a formal proposal, not an activated rule, per bip360.org. The current draft does not itself introduce a post-quantum signature; it restructures the output and defers the new cryptography to follow-on work.

The harder problem is the coins already exposed. BIP-361, whose authors include Jameson Lopp, tackles migration with a phased sunset of legacy and Schnorr signatures: a first phase (on the order of three years after activation) that blocks new sends to quantum-vulnerable address types, and a second (around two years later) that stops honoring vulnerable-signature spends altogether, with a zero-knowledge rescue mechanism so legitimate owners can prove seed ownership and recover otherwise frozen funds, per the bitcoin/bips repository. Lopp has been careful about the ethics: “No one can be forced to migrate their coins,” he has written, but “no one can be forced to accept coins from quantum-vulnerable signature schemes either.”

None of this is close to activation, and each piece revives old fights. Post-quantum signatures are far larger than Schnorr, which reopens the block-size debate, and any freeze-and-rescue design has to answer for coins whose owners are gone (including, most sensitively, Satoshi’s). The comparison with Ethereum is once again pointed: Ethereum’s foundation can push a top-down roadmap toward post-quantum cryptography, while Bitcoin has to earn the same change through the slow, adversarial consensus process that just chewed up BIP-110.

What the SEC Can and Can’t Touch

For US readers watching this from a regulatory angle, one point matters more than any other: none of it is something the Securities and Exchange Commission approves or blocks. A soft fork like Taproot, or a future activation of BIP-360, is a change to Bitcoin’s consensus rules, and the consensus layer sits outside the SEC’s remit. No regulator green-lit Taproot in 2021, and none can veto a post-quantum upgrade in 2028.

Where the rules bite is downstream of the protocol. Custodians and spot Bitcoin ETF issuers have to write policies for how they would handle a contentious fork or an airdropped asset like the eCash coins due later this month, and those policies are supervised. The Internal Revenue Service treats forked or airdropped coins received by a taxpayer as ordinary income at the time of receipt under its 2019 guidance, so the eCash drop is a taxable event for US holders regardless of whether they wanted it. And any token issued on Bitcoin through Taproot Assets or a covenant-based system inherits the same securities-law questions that hang over tokens everywhere, which is where enforcement actually happens, as we lay out in our field guide to how SEC crypto enforcement works in 2026.

The clean way to think about it: regulators supervise the on-ramps, the custodians and the issuers, not the signature scheme. Taproot changed what Bitcoin can express; it did not change who answers to Washington.

Price, the Quantum Discount, and What Comes Next

Does any of this reach the price? At around $64,000, roughly 49% below its October 2025 high of $126,080, Bitcoin trades on macro liquidity and ETF flows far more than on cryptography, per CoinGecko. But the quantum debate has started to seep into valuation talk. Capriole Investments founder Charles Edwards has floated a Quantum Discount Factor, arguing that unresolved quantum risk should shave something off Bitcoin’s fair value until the migration path is credible, and has warned of a much lower price scenario later this decade if the issue festers. Whether you buy that framing or file it under tail-risk theater, it is now part of the analyst conversation, one strand of the broader picture we chart in our guide to reading the 2026 analyst target map.

The nearer-term outlook for Taproot is less dramatic and more encouraging. The five-year anniversary of activation arrives in November 2026 with the upgrade doing exactly what it was designed to do: making complex Bitcoin cheaper and more private, quietly, one wallet and one protocol at a time. Adoption should keep grinding up as bc1p support fills in. Stablecoins and Lightning give it a demand source that does not depend on inscription manias. Covenants are stalled at the consensus layer but routed around by BitVM2 rollups like Citrea. And the quantum question, uncomfortable as it is, has forced Bitcoin to start planning its biggest cryptographic change ever while it still has years to do it carefully.

The BIP-110 fight showed that Bitcoin will not casually restrict what Taproot enables. The Google research showed that Taproot’s openness carries a long-dated cost. Reconciling those two facts, keeping the expressiveness while closing the exposure, is the work of the next five years.

Frequently Asked Questions

What is Bitcoin Taproot in simple terms?

Taproot is a 2021 Bitcoin upgrade (BIPs 340, 341 and 342) that added Schnorr signatures and MAST. It lets complex transactions such as multisig and Lightning channels look like ordinary single-signature payments, making them cheaper and more private. Taproot addresses start with bc1p.

When did Taproot activate and how widely is it used?

Taproot activated at block 709,632 on November 14, 2021. Its transaction share peaked above 40% in 2024 during the Ordinals and Runes boom and has since settled to roughly a fifth of Bitcoin activity, with native SegWit still the most common format.

Why is Taproot considered vulnerable to quantum computers?

Taproot outputs reveal the full public key on-chain by default, unlike legacy and SegWit addresses that hide it behind a hash until first spend. A future quantum computer able to break elliptic-curve cryptography could target that exposed key. Proposals BIP-360 and BIP-361 aim to move Bitcoin to quantum-resistant schemes.

What was BIP-110 and why did it fail?

BIP-110 was a temporary soft fork meant to restrict data-carrying transactions, mostly Ordinals inscriptions, that Taproot made cheap. It needed broad miner support but drew only about 2.53% signaling. In August 2026 a minority chain enforcing it split off, mined two blocks and stalled, while about 99.85% of hashpower stayed on the main chain.

Is Taproot regulated by the SEC?

No. Taproot is a change to Bitcoin’s consensus rules, which sit outside the SEC’s authority. Regulators supervise custodians, exchanges and token issuers, not the protocol itself. Downstream effects, such as how ETFs handle forks or how the IRS taxes airdropped coins, are where oversight applies.

Marcus Okafor is a senior editor at HOGE Wire covering Bitcoin’s base layer, mining, and protocol development.