Wrapped Bitcoin Is a Bridge: WBTC, cbBTC, tBTC in 2026

In the seven days before this article went out, three cross-chain bridges broke in almost the same way. Cysic lost control of its minting path on 17 October, and an attacker minted 10 million unauthorized CYS tokens on BNB Smart Chain. On 22 October the WAX multichain bridge released tokens on Solana and Ethereum that were never paid for on the WAX side, roughly 75,000 to 90,000 dollars across five transactions. On 23 October Initia halted its entire chain after a script-caching flaw in its MoveVM let an attacker mint unbacked bridged assets, about 2.7 million dollars of which moved to Ethereum, according to a roundup of the week by security firm Shattered Labs. Then, on the morning this piece published, the ZKLink Nova bridge was drained of roughly 780,000 dollars in ETH and tokens through a forged batch root, leaving its escrow nearly empty, per Phemex, citing the security firm Blockaid.

Bitcoin, meanwhile, was trading around 81,000 dollars after the Federal Reserve held its benchmark rate at 3.75 to 4.00 percent on 28 October, as CoinDesk reported live. None of those four broken bridges touched Bitcoin directly. But the single largest bridge exposure most crypto users hold is not a protocol they think of as a bridge at all. It is wrapped Bitcoin: the roughly 18 billion dollars of BTC that now lives as tokens on Ethereum, Solana, Base and other chains. Every unit of it crossed a bridge to get there, and the kind of bridge it crossed decides how safe it is.

Why a Wrapped Bitcoin Token Is a Bridge

Bitcoin cannot natively run inside an Ethereum smart contract or a Solana program. Its scripting language is deliberately limited, and its chain has no idea what is happening anywhere else. So to use BTC as collateral in a lending market, as a trading pair on a perpetual-futures venue, or as liquidity in an automated market maker, somebody has to represent it on the other chain. That representation is what wrapped Bitcoin means: your real BTC is locked or custodied on one side, and a matching token, an IOU, is minted on the other.

Lock on one chain, mint on another: that is the textbook definition of a bridge. A unit of wrapped Bitcoin is worth a real bitcoin only if the thing holding the underlying BTC stays honest and solvent, and only if the contract that mints the token refuses to issue more than the BTC actually in the vault. The token is a claim, not the coin. When people argue about whether a given wrapped Bitcoin is safe, they are really arguing about the bridge sitting underneath it.

Roughly 18 billion dollars of Bitcoin now exists in this tokenized form, per on-chain trackers, and that is still only a little over 1 percent of all BTC. The other 99 percent sits on Bitcoin’s base layer, in exchange omnibus accounts, or in cold storage, a world away from the smart-contract risk of the wrappers and secured instead by Bitcoin’s miners and the economics explored in our look at the shakeout among Bitcoin mining pools. Native issuance on Bitcoin itself, such as the Runes tokens that live directly on Layer 1, never crosses a bridge at all. The moment BTC leaves that base layer to do anything programmable, a bridge appears.

Vitalik Buterin warned about this as early as 2022, arguing that there are “fundamental limits to the security of bridges” once value hops across multiple “zones of sovereignty,” a point Cointelegraph reported at the time. Wrapped Bitcoin is exactly that kind of hop: the asset lives under one chain’s rules while its claim trades under another’s, and the bridge is the only thing reconciling the two.

The Two Things Every Wrapped BTC Has to Get Right

Strip away the branding and every wrapped Bitcoin has to get exactly two things right. First, custody: the real BTC has to be held somewhere, by someone or something that cannot run off with it or lose the keys. Second, accounting: the contract that mints the token has to issue one unit only when one real bitcoin arrives, and burn one unit only when one real bitcoin leaves. Break the first and the backing disappears. Break the second and the supply inflates past the backing. Either way the peg dies.

It helps to picture the full cycle. To mint, you send real BTC to the bridge’s custody and the contract issues you an equal amount of the wrapped token; to redeem, you burn the token and the custody releases the underlying coin. As long as that loop stays one-to-one, the wrapper trades at par with Bitcoin because anyone can arbitrage a gap by minting or redeeming against it. The peg is not a law of nature; it is a standing promise that redemption will always be honored one-for-one. Every attack on a wrapped Bitcoin is, at bottom, an attack on that promise, either by stealing the backing or by minting claims the backing cannot cover.

Those two failure modes map almost perfectly onto how bridges have actually broken. The costliest incidents of the last two years were not clever contract exploits; they were compromised keys and forged messages. TRM Labs found that in the first half of 2026, infrastructure, key and operational compromises made up only about 15 percent of incidents but roughly 76 percent of the value stolen, in its mid-year hacks report. The lesson the industry keeps relearning is that the code is rarely the weak point.

Ronghui Gu, co-founder of the audit firm CertiK, put it bluntly: “A protocol can pass a flawless code audit and still lose millions because of a compromised admin key,” he told Forbes. For wrapped Bitcoin that warning cuts both ways: you can read every line of the token contract and still be wiped out if the entity holding the BTC is captured, or if the mint can be tricked into issuing coins that nothing backs. So the useful question is never just “is it audited.” It is “who holds the BTC, and what would it take to make the mint lie.”

The Custody Spectrum, From One Company to No Company

The wrapped-Bitcoin market sorts itself along a single axis: how many parties you have to trust, and how hard it is to corrupt them. At one end sits a single regulated company holding every coin. In the middle sit federations and multi-party custodians, where a threshold of signers has to agree. At the far end sit trust-minimized designs where cryptography splits the key across a large, shifting set of operators so that no individual, and no small group, can move the Bitcoin alone. Every product below is a bridge; they just differ on who is standing on it.

The largest wrapped Bitcoin, WBTC, holds about 116,000 BTC, worth roughly 9 billion dollars at an 81,000-dollar price, per DeFiLlama. Coinbase’s cbBTC is a close second at about 99,000 BTC, or roughly 8 billion dollars, per CoinGecko. Everything else, the staked and restaked and trust-minimized wrappers, is far smaller. The table below uses token counts, which do not move with the BTC price, drawn from a mid-September 2026 tokenized-Bitcoin supply tracker.

Note what the sizes tell you. Nearly all tokenized Bitcoin sits in the two most custodial products. The trust-minimized option, tBTC, is a rounding error by comparison. That is the market quietly admitting it prefers a known company it can sue to a cryptographic design it has to understand, a preference that holds right up until the company does something the market did not expect.

WBTC and the Custodian That Changed Hands

WBTC is the oldest and still the largest wrapped Bitcoin, live since 2019 and custodied for years by BitGo, a US-based institutional custodian. For most of its life the pitch was simple: a known, regulated custodian holds the BTC, you hold the token. Then the custody changed hands, and WBTC became the clearest case study in why the custodian matters more than the code.

In August 2024, BitGo announced it would move WBTC from sole US control into a multi-jurisdictional arrangement with BiT Global, a Hong Kong entity tied to Justin Sun and the TRON ecosystem, with the three custodial multisig keys to be split between BitGo, BiT Global and a BiT Global affiliate in Singapore. The market reacted fast. Coinbase said on 19 November 2024 that it would delist WBTC the following month, citing its listing standards, and launched its own wrapped Bitcoin, cbBTC, as The Block reported. BiT Global sued Coinbase for anticompetitive conduct; a judge let the delisting proceed, and BiT Global later dropped the suit with prejudice. Sky, the lending protocol formerly known as MakerDAO, moved to offboard WBTC as collateral.

Nothing about the WBTC contract changed during any of this. The Solidity was the same before and after. What changed was who held the keys to the real Bitcoin, which is the entire security model of the token. It is the same lesson DeFi keeps learning about control: the risk is in who can move the assets, a dynamic we traced in our reporting on governance capture and who really owns a protocol. A wrapped Bitcoin inherits the trustworthiness of its custodian, full stop, and custodians can be acquired, pressured or replaced.

WBTC has also been quietly re-platforming the bridge that moves it between chains. In August 2026 BitGo made Chainlink’s CCIP the sole cross-chain provider for more than 7 billion dollars of WBTC, part of a roughly 15-billion-dollar migration of tokenized assets away from LayerZero toward CCIP, as CoinDesk reported. Even the oldest wrapper is treating its cross-chain messaging as a security decision worth re-making.

cbBTC: A Regulated Custodian, and a Freeze Switch

Coinbase’s cbBTC took the opposite lesson from the WBTC saga and leaned into it. Instead of spreading custody across jurisdictions, Coinbase Custody, the institutional arm of a US-listed public company, holds all the underlying BTC one-for-one. For Coinbase users the experience is seamless: deposit BTC, receive cbBTC on a supported chain, with no minting fee, and redeem by routing back through a Coinbase account. Launched in September 2024, cbBTC is now live on Ethereum, Base, Solana and Arbitrum, and it has grown to roughly 99,000 BTC, making it the second-largest wrapped Bitcoin.

The strength and the weakness are the same fact. A regulated, audited public company is about as solid a counterparty as this market offers, which is exactly why cbBTC grew so fast. But it is a single issuer with a freeze switch. There is no permissionless, self-service redemption for someone who is not a Coinbase customer; the path back to real BTC runs through one company’s compliance desk. If Coinbase were compromised, compelled by a court, or simply decided to freeze an address, every cbBTC holder is exposed to that one decision at the same moment. You are not trusting math or a federation. You are trusting a balance sheet and a legal department, and betting they stay aligned with you.

tBTC: Replacing the Custodian With Cryptography

tBTC, from Threshold Network, is the market’s main attempt to delete the custodian entirely. There is no company holding the BTC. Instead the signing key that controls deposited Bitcoin is split, using threshold cryptography, across a large and randomly selected set of node operators, reported to be around 100, a majority of which must cooperate to move any coin. No single operator, and no small cabal, ever holds the whole key or can act alone. You mint by sending BTC to a protocol-generated address and, once it confirms, the contract issues tBTC; you redeem by burning tBTC and naming a Bitcoin address for the signers to pay out. By 2026 the protocol had added gasless minting and set a mint-and-redeem fee of around 20 basis points under a governance vote, as a 2026 review of wrapped-Bitcoin options documented.

The honest framing is trust-minimized, not trustless. You remove the single-custodian failure mode, and in exchange you take on smart-contract risk and operator-set risk: the circuits and the signer selection have to be sound, and enough operators have to stay honest and online. The market’s verdict shows in the size. At roughly 4,300 BTC, a few hundred million dollars, tBTC is a fraction of WBTC or cbBTC, which means thinner liquidity if you need to exit in a hurry and more conservative treatment in lending markets that accept it. The cryptography is the most elegant answer to the custody problem, and it is also the least used, which tells you how much the market actually values convenience and depth over decentralization.

The Yield-Bearing Wrappers: Lombard, FBTC and the New Wave

A newer cohort of wrappers does not just hold your Bitcoin; it puts the Bitcoin to work. Lombard’s LBTC, at roughly 10,000 BTC, stakes the underlying coin through the Babylon protocol and pays a yield, with custody run by a federation of security partners. Function’s FBTC, around 8,400 BTC, uses multi-party-computation custody plus on-chain attestations. Ether.fi’s eBTC is Babylon-secured and restaked. Circle, meanwhile, has introduced cirBTC, a small, newer entrant that emphasizes native, continuously attested on-chain reserves rather than a black-box balance sheet.

Yield is not free; it is a second trust layer stacked on the first. With a plain custodial wrapper the only question is whether the custodian holds your coin. With a staked or restaked wrapper you also have to ask what the coin is doing while wrapped, and what can go wrong there: a slashing event, a bug in the staking protocol, a failure in the federation that manages the position. Each additional layer is another bridge, another verifier, another thing that has to stay honest for your claim to redeem one-for-one. The marketing calls it Bitcoin that earns. The security engineer calls it more surface area.

Liquid and the 320-Million-Dollar Lesson in Federated Pegs

If you want proof that a reputable federation is still a bridge, look at Liquid. Blockstream’s Liquid Network is an eight-year-old Bitcoin sidechain whose token, L-BTC, is backed one-for-one by BTC held in an 11-of-15 federation multisig: fifteen known functionaries, eleven of whom must sign to move the coins. On paper that is a strong bar. On 6 September 2026 it failed anyway. Roughly 4,000 BTC, about 320 million dollars at the time, left the federation wallet, around 95 percent of the reserve, as CoinDesk reported.

The root cause was not stolen keys. It was a cache-key collision in the range-proof verification inside Elements, Liquid’s software: an attacker got a valid proof cached, then submitted different, invalid data that resolved to the same cached result, so the honest federation signed off on a forged mint. Chainalysis described the trick as getting valid data verified and cached, then pointing invalid data at the same cached result. The eleven honest signers did their jobs. They just verified something that was false. The high multisig threshold was worthless because the thing being checked was a lie.

This was the largest single bridge loss of 2026, ahead of the roughly 292-million-dollar KelpDAO exploit in April; the all-time record remains the roughly 625-million-dollar Ronin bridge hack of 2022. The self-described white-hat movers returned about 3,400 BTC and kept around 600 as a self-declared bounty, and Liquid resumed block production on 10 September after Blockstream patched the bug, according to News.Bitcoin.com. Blockstream chief executive Adam Back urged L-BTC holders not to “panic sell OTC” and said the peg with Bitcoin “will be covered” one-for-one. The reassurance is the point: a federation is a counterparty, and when it fails, holders are left waiting on a company’s promise to make the peg whole.

When the Bridge Believes a Lie

Liquid and the October wave share one failure mode, and it is the one that now dominates bridge losses. None of these bridges lost their keys. Each was fed false information and accepted it. ZKLink Nova accepted a forged batch root. Initia’s MoveVM script cache bypassed its own security checks. WAX released tokens it was never paid for. Liquid reused a cached proof. In every case the authorization machinery worked exactly as written; it simply authorized something untrue.

Ben Fisch, chief executive of Espresso Systems, made the distinction precisely after the KelpDAO bridge exploit earlier in the year. The bridge, he said, “worked as designed”; it had “just believed the wrong information,” as he told CoinDesk. A signature check confirms who sent a message, not that the message is true. For wrapped Bitcoin, that is why the minting logic matters as much as the custody. If the mint can be tricked into issuing unbacked tokens, the peg breaks even though every real bitcoin is still sitting in the vault. The table below maps the trust models to the way each one actually fails.

How the Industry Is Hardening Bitcoin Bridges

The fixes fall into four buckets, and the better wrappers are adopting several at once. The first is transparent, native reserves: instead of asking you to trust a quarterly attestation, newer designs publish a continuously verifiable, on-chain link between tokens minted and BTC held, the approach Circle markets for cirBTC and the same burn-and-mint logic Circle uses for native USDC through its Cross-Chain Transfer Protocol, which issues the asset natively on each chain rather than pooling a wrapped IOU. The second is harder cross-chain messaging, such as WBTC’s move to Chainlink CCIP with its separate risk-management network, so that moving the wrapper between chains does not rely on a single committee.

The third is rate limiting. Token standards like xERC-20 let an issuer cap how much any one bridge can mint in a given window, so that even a bridge that is tricked into issuing fake tokens can only issue so many before the cap trips, turning a catastrophic unbacked mint into a contained one. The fourth is cryptographic verification: light clients and zero-knowledge proofs that let a destination chain check a source chain’s state with math rather than taking a committee’s word for it. None of these is a silver bullet, and stacking them is the point; the industry now calls this defense in depth, which is a polite way of saying no single layer should be trusted to hold alone.

For wrapped Bitcoin specifically, proof of reserves is the defense that fits the threat. An unbacked mint, the WAX and Initia and ZKLink failure, is exactly the thing a live, on-chain reserve check is built to catch. It does not stop the first fraudulent token from being issued, but it shrinks the window between the lie and its discovery from weeks to minutes, and in a bank run on a peg, minutes are most of the game.

The Institutional Version Looks Nothing Like Yours

There is a quiet split forming in how Bitcoin crosses chains. Retail users reach for whatever wrapped BTC is liquid on the chain they are using, and they inherit whatever trust model that token happens to carry. Institutions moving tokenized treasuries, funds and large BTC positions have started to treat the bridge as a procurement decision, and they are buying very different infrastructure. The roughly 15-billion-dollar migration of tokenized value from LayerZero to Chainlink’s CCIP in 2026, which swept up BitGo’s WBTC alongside tokenized funds and stablecoins, was not about speed or fees. It was a flight to a messaging layer with a separate risk-management network and configurable rate limits, chosen by desks that answer to risk committees rather than to a Discord.

The same logic is reshaping the wrappers themselves. Institutional-grade products lean on named, regulated custodians, continuous proof-of-reserves attestations, and native issuance that avoids pooling a wrapped IOU anyone can drain. The irony is that this is the opposite of the permissionless ideal: the safest-feeling Bitcoin bridges in 2026 are the ones with the most identifiable, suable counterparty on the other end. That is a reasonable trade for a treasurer and a strange one for a cypherpunk, and it is worth being honest about which of the two you are when you choose a wrapper. The honeypot does not shrink because the holder wears a suit; a larger pool of tokenized Bitcoin is simply a larger prize, which is why the hardening above matters most exactly where the value concentrates.

What a Spot Bitcoin ETF Does Not Give You

US readers have a cleaner, regulated way to hold Bitcoin exposure: the spot Bitcoin exchange-traded funds that the SEC approved in 2024. But an ETF share is not on-chain BTC. It is an equity-like claim held through a broker, and you cannot post it as collateral on Aave, pair it on a decentralized exchange, or do anything programmable with it. The entire reason wrapped Bitcoin exists is that ETFs solved the access problem and left the composability problem untouched. Wrapped BTC is how BTC becomes programmable, and it largely sits outside the ETF’s investor-protection perimeter.

The regulatory picture around what wrapped Bitcoin is has actually gotten clearer. In March 2026 the SEC and CFTC issued a joint interpretation treating most crypto assets, Bitcoin included, as digital commodities rather than securities, as Decrypt reported. That tells you how a wrapped bitcoin is likely classified. It does not tell you who makes you whole after a bridge hack, and the answer is usually no one. There is no FDIC, no SIPC, no ETF authorized participant to call. When a federation signs a forged mint or a custodian is compromised, recovery is a white-hat negotiation or nothing, and funds laundered through sanctioned mixers, the territory of OFAC’s Tornado Cash designations and the Roman Storm prosecution, rarely come back. A regulator can punish a custodian after the fact. It cannot un-mint a fake token or reverse a settled Bitcoin transaction.

How to Read a Wrapped Bitcoin Before You Hold It

You do not need to audit Solidity to evaluate a wrapped Bitcoin. You need to answer a short list of questions about the bridge underneath it, and then decide whether the trust model is one you are comfortable holding. Before you mint, bridge into, or accept a wrapped BTC as collateral, work through these.

• Who holds the underlying BTC, and can they freeze or seize it? A single custodian is convenient and freezable; a federation spreads the risk; a threshold scheme removes the single controller.
• Is there live, verifiable proof of reserves, or just a periodic attestation? Continuous on-chain proof catches an unbacked mint fast; a quarterly PDF does not.
• Who can mint, and what would it take to trick the mint into issuing unbacked tokens? This is the failure mode behind most of 2026’s losses.
• How do you redeem, and can anyone redeem, or only the issuer’s own customers? A redemption path that runs through one company is a freeze point.
• Which chain and which bridge does this version ride? A token is only as safe as the weakest chain and messaging layer it is issued across.
• How deep is the liquidity if you need to exit fast during a depeg scare? Thin markets turn a scare into a loss.

The point of the checklist is not to steer you to one product. It is to make the trade explicit. The convenience and depth of the big custodial wrappers is real, and so is their single-issuer risk; the decentralization of tBTC is real, and so is its thinner liquidity. The table below is a quick way to sort the green flags from the red ones.

The bottom line is unglamorous but worth repeating every time a bridge breaks: wrapped Bitcoin is not free, programmable BTC. It is a bridge position with a counterparty, priced at par only as long as the bridge holds. Choose the trust model on purpose, size the position to the risk, and never assume the token in your wallet is the same thing as a coin on Bitcoin’s base layer. It is a claim on one, and the distance between the claim and the coin is exactly the bridge.

Frequently Asked Questions

Is wrapped Bitcoin safe?

Wrapped Bitcoin is as safe as the bridge and custodian behind it, not as safe as Bitcoin itself. Each unit is a claim on real BTC held by a custodian, a federation or a cryptographic signer set, and your risk is whatever can compromise that holder or trick its minting contract. A spot Bitcoin ETF or self-custodied BTC carries no such bridge risk.

What is the difference between WBTC, cbBTC and tBTC?

WBTC is held by a custodian multisig split between BitGo and BiT Global; cbBTC is held one-for-one by Coinbase Custody, a single regulated company with a freeze capability; and tBTC is trust-minimized, with the key split across around 100 node operators so no single party controls the Bitcoin. WBTC and cbBTC are far larger and more liquid, while tBTC removes the single custodian at the cost of smaller size.

Why did Coinbase delist WBTC?

Coinbase announced in November 2024 that it would delist WBTC, citing its listing standards, after custodian BitGo moved WBTC into a multi-jurisdictional arrangement with BiT Global, an entity tied to Justin Sun. Coinbase launched its own wrapped Bitcoin, cbBTC, around the same time. BiT Global sued over the decision and later dropped the case.

Can wrapped Bitcoin lose its peg?

Yes. A wrapped Bitcoin can depeg if the custodian loses or cannot return the underlying BTC, or if the bridge is tricked into minting tokens that nothing backs. The Liquid Network lost about 4,000 BTC in September 2026 when its federation signed a forged mint, and several October 2026 bridges released unbacked tokens. The peg holds only as long as backing and honest accounting both hold.

Is a spot Bitcoin ETF safer than wrapped Bitcoin?

For simply holding Bitcoin exposure, a regulated spot ETF avoids bridge and custodian-peg risk, but it gives you an equity-like share you cannot use in DeFi. Wrapped Bitcoin is what makes BTC programmable on other chains, and that utility comes with the bridge risk an ETF does not have. They solve different problems.

By Yuki Tanaka, HOGE Wire DeFi and on-chain desk.