📖 Why tokens are moved between blockchains
A crypto bridge is a protocol that moves tokens and messages between two blockchains. In a lock-and-mint model, the token is locked on the source chain and a wrapped token is issued on the destination chain. In a liquidity-pool model, the token is sent into a pool on the source chain and the same token is released from a pre-funded pool on the destination chain. DeFi, or decentralized finance, means financial services built on smart contracts without banks.
Wrapped token is a version of the original asset on the destination chain issued by a smart contract. This token is backed by a reserve: the original asset is locked in the bridge contract on another chain. A wrapped token can be redeemed back into the original asset if the reserve is intact and the bridge correctly verifies cross-chain messages.
A crypto bridge controls two resources: the reserve of locked tokens in the contract and the right to mint or release an asset on the destination chain. If the contract is exploited, signers are compromised or cross-chain message verification breaks, one of two things happens: the reserve is drained from the bridge, or unbacked wrapped tokens are minted and sold on a DEX.
Cross-chain infrastructure risks overlap with common DeFi threats: phishing, malicious approvals and key compromise. A threat map and a breakdown of wallet protection models are covered in “DeFi security: threat map, cases and protection”.
In the interface, a cross-chain transfer looks like a single operation. At the protocol level, the transfer has two stages: an event is recorded and confirmed on the source chain, then the asset is separately released or minted on the destination chain.
Bridge risk depends on how the bridge moves the asset and who confirms cross-chain messages.
Crypto bridges: basic overview and purpose
Crypto bridges move tokens and messages between blockchains. Multichain means working with assets across several networks; DeFi means financial services on smart contracts without banking intermediaries.
A crypto bridge performs a transfer in two ways: (1) the asset is locked on the source chain and a wrapped token is minted on the destination chain; (2) the asset moves through liquidity pools, where release on the destination chain comes from a pre-funded reserve.
Bridges are used to move stablecoins to a network with lower fees, transfer an asset to a network with the needed application, or support protocols that operate across several blockchains.
Bridge risk depends on where the reserve is held and what mechanism confirms the release or minting of the asset on the destination chain.
The same operation in the interface can rely on different mechanics and different sources of trust.
Types of crypto bridges: transfer mechanics and trust model
Two bridges can look identical in the interface but lose funds for different reasons. The comparison comes down to two questions: what operation makes the asset available on the destination chain, and who confirms the event on the source chain.
By token transfer mechanics
- Lock-and-Mint. The token is locked in the bridge contract on the source chain, and a wrapped token is issued on the destination chain. During a reverse transfer, the wrapped token is burned and the original token is unlocked.
- Burn-and-Mint. The token is destroyed, or burned, on one chain and minted on another. This model is used when issuance and burning are governed by unified rules of the issuer or protocol.
- Liquidity pools, or lock-and-unlock in application form. Each chain has a pre-positioned reserve of the token. The user sends the token into a pool on the source chain and receives the same token from the pool on the destination chain without minting a wrapped version.
Atomic swaps (HTLC) use HTLC contracts, or Hashed Timelock Contracts: the asset is released only when a secret whose hash is known in advance is revealed, and only before the deadline. HTLC limitations are tied to support for specific networks and the need to coordinate timelock parameters across two contracts.
By security model
- Custodial. Asset release is controlled by a company or a limited group of operators. The key risks are key compromise and service shutdown.
- Non-custodial. Asset release is confirmed by a distributed group of validators or signers. An attack requires control over the signature threshold or the stake.
- Native bridges. Cross-chain interaction is built into the blockchain protocol or a related ecosystem. Trust relies on the security of the base network.
- Universal bridges. One protocol connects many networks through a shared set of contracts and a message verification mechanism. A fault in the shared infrastructure affects many routes.
Moving BTC into DeFi often uses wrapped versions, where the collateral custodian and the wrapper redemption procedure are critical. An example of the product and its related infrastructure dependency is described in “Yield Basis: Bitcoin yield without IL”.
The reliability of a crypto bridge is defined by the combination of token transfer mechanics and the cross-chain message confirmation model.
Losses in crypto bridges arise at the points where the bridge stores reserves and where it confirms events between networks.
Crypto bridge risks: where and why losses occur
A crypto bridge consists of several parts. The point of failure can be the reserve contract, the cross-chain message confirmation mechanism or the governance layer that can stop or change how the bridge operates.
Errors in bridge code
- How it happens: the bridge accepts a transfer as confirmed even though the event on the source chain is incorrect or forged.
- Why it is dangerous: locked tokens can be withdrawn from the reserve contract, or tokens without backing can appear on the destination chain.
- What to check: whether public smart-contract audits exist, whether a bug bounty is active and who approves contract upgrades.
Wrapped token risk
- What it means: a wrapped token holds its price only while the original asset is actually locked in the bridge reserve.
- When the problem appears: if the bridge stops or loses its reserve, redeeming the wrapped token into the original asset becomes impossible or happens at a discount.
- What to pay attention to: the reserve address and custody model, the redemption procedure and the wrapper liquidity available for exit.
Transfer and confirmation of messages
- How it works: the bridge records an event on the source chain and sends a message to the destination chain, where release or minting is performed.
- Where risk appears: with false confirmation or delivery failure, the bridge can release an asset without a valid event on the source chain.
- What matters: how many independent parties confirm the message and whether transfers can be paused when a failure is detected.
Governance and human factor
- How decisions are made: bridge parameters are changed through admin keys, a multisig or validators.
- Why this is a risk: key compromise or an upgrade mistake can stop the bridge or give an attacker the right to mint and release assets.
- What to consider: the signature threshold for changes, the presence of a timelock and limits on release volumes.
A crypto bridge assessment comes down to three checks: where the reserve is held, who confirms cross-chain messages and who can change the bridge rules.
The same interface action can rely on different liquidity sources and different confirmation schemes.
Overview of key crypto bridges: examples, strengths and limitations
A cross-chain transfer can be executed by releasing funds from a liquidity pool, minting a wrapped token, or delivering a message that triggers a later action on the destination chain.
Stargate
GlobalMoves stablecoins between EVM networks through liquidity pools. On the source chain, the transaction sends the token into a pool. The transfer message is delivered through LayerZero. On the destination chain, the token is released from a pool as the same asset, without minting a wrapped version.
✅ Pros
- Release on the destination chain comes from a liquidity pool without a wrapped token.
- Routes are focused on liquid stablecoins, where it is easier to exit without major slippage.
- Risk depends on message delivery and on pool balances on the destination chain.
❌ Cons
- The asset set is limited: a pool is required for every token and every network.
- Message confirmation depends on the LayerZero configuration, including oracle and relayer.
- A liquidity shortage in the pool limits the release volume on the destination chain.
The transfer relies on message transport and on available reserves in destination-chain pools.
LayerZero
GlobalA protocol for delivering messages between blockchains. Event confirmation is built on an “oracle + relayer” pair: the oracle provides state data, the relayer delivers the payload, and the application defines validation rules and trust thresholds.
✅ Pros
- Supports cross-chain messages and contract calls, not only token transfers.
- One message transport is used across many networks through applications built on top of the protocol.
- Resilience can be improved by separating data sources and delivery sources.
❌ Cons
- If the oracle and relayer are controlled by one party, the risk of false confirmation increases.
- Confirmation rules are set by the application and may not be visible in the interface.
- Route risk depends on the selected sources and on the validation thresholds in the application.
Transfer security depends on the oracle, relayer and verification rules configured in the specific application.
Synapse
GlobalThe protocol combines cross-chain transfer and token exchange: an asset can be moved to another network and swapped into another token along the route. The transfer is confirmed with a delay. During this window, special participants can challenge an invalid operation. If a participant confirms a fraudulent operation, their collateral is slashed, making false confirmations economically unattractive.
✅ Pros
- Transfer and swap can be completed in one operation.
- Common EVM networks and L2s are supported.
- Collateral from confirming participants creates an economic barrier against false confirmations.
❌ Cons
- The challenge window requires active watchers to stop anomalies.
- Risk depends on collateral distribution and the behavior of confirming participants.
- On rare pairs, liquidity may be insufficient for large volumes.
Transfer and swap happen together, while risk is tied to missing an anomaly during the challenge window.
Axelar
GlobalConfirms cross-chain messages through its own PoS network. PoS, or Proof of Stake, uses validators with stake that can be penalized for rule violations. Release or minting on the destination chain happens after Axelar validators sign the message.
✅ Pros
- Message confirmation is distributed across validators of the PoS network.
- One protocol connects EVM and Cosmos for cross-chain messages.
- Supports message passing and contract-call scenarios.
❌ Cons
- Delay includes finality of the source chain and processing inside the PoS network.
- Risk depends on stake distribution and validator resilience.
- The fee includes operations on several networks and message execution.
Transfer depends on source-chain finality and on the Axelar validator signature threshold.
Wormhole
GlobalWhen a token is transferred, the asset is locked on one chain and its wrapped version is minted on another. Minting happens after the lock is verified by a group of independent nodes. The wrapped token appears on the destination chain only after a required number of those nodes agree.
✅ Pros
- Supports EVM networks and alternative ecosystems, including non-EVM.
- Wrapped token minting happens after threshold confirmation.
- Supports cross-chain messages for applications.
❌ Cons
- The guardian model remains a limited signer set.
- Wrapped tokens depend on reserve safety and redemption availability.
- Historical incidents increase the importance of limits and monitoring.
The transfer depends on the guardian signature threshold and on the reserve that backs the wrapped asset remaining intact.
Celer cBridge
GlobalExecutes transfers through liquidity pools, while event confirmation and release rely on the State Guardian Network, or SGN. SGN watches events on supported networks and signs permission to release funds from pools on the destination chain.
✅ Pros
- Release through liquidity pools reduces dependence on wrapped tokens in typical routes.
- Threshold signatures in SGN reduce the risk of a single admin key.
- Integrations with interfaces and aggregators make routes easier to access.
❌ Cons
- Confirmation risk depends on the composition and distribution of SGN validators.
- A liquidity shortage in the destination pool limits release for specific tokens.
- Economic incentives are partly tied to the CELR token.
The transfer depends on SGN signatures and on reserves in liquidity pools for the specific tokens and networks.
Bridge selection comes down to two checks: who confirms messages and where liquidity for release on the destination chain comes from.
The comparison table helps connect the transfer mechanism with where the trusted surface is located.
Comparison of popular crypto bridges: mechanism and trusted surface
In the table, “release delay” means the interval between transaction confirmation on the source chain and token appearance on the destination chain. The value depends on source-chain finality and on message processing by the bridge mechanism.
| 🌉 Bridge | 🌐 Supported networks | 🛠 Mechanism | 🔐 Security model | ⏱ Release delay |
|---|---|---|---|---|
| Stargate | EVM networks (L1/L2) | Liquidity pools | Message transfer through LayerZero | Minutes |
| LayerZero | EVM + non-EVM (through applications) | Messages between chains | Oracle + independent relayer | Minutes |
| Synapse | EVM networks (L1/L2) | Liquidity + swap | Optimistic verification with challenge | Minutes |
| Axelar | EVM + Cosmos | Lock-and-Mint | Own PoS validator network | Minutes |
| Wormhole | EVM + Solana + non-EVM | Lock-and-Mint | Threshold guardian signatures | Minutes |
| Celer cBridge | EVM networks (L1/L2) | Liquidity pools | State Guardian Network (SGN) | Minutes |
| Multichain | Many networks | Lock-and-Mint | MPC nodes | Minutes |
| Hop | L2 networks (L2↔L1 routes) | Liquidity pools | Bonders with collateral | Seconds to minutes |
Liquidity pools reduce dependence on wrapped tokens, but they require reserves on the destination chain. Lock-and-mint depends on reserve safety and on the robustness of the signature mechanism that authorizes wrapped token issuance.
Bridge security checks are about reducing the number of trusted components and making it transparent who confirms cross-chain messages and how they do it.
Crypto bridge protection is built around the reserve, message confirmation and upgrade control.
Crypto bridge security: governance, audits, limits and proofs
Resilient architectures separate the reserve contract from message confirmation logic. The reserve contract is limited to locking and releasing funds. Message transport and verification are moved into separate modules.
- Multisignatures and distributed governance
- Contract upgrades and limit changes are confirmed by a multisig scheme, not by a single key.
- A signature threshold, for example 5 of 8, reduces the risk that one participant compromises governance.
- A timelock delays the application of upgrades and reduces the risk of instant logic replacement.
- Audits and bug bounties
- An audit checks message verification, access rights and balance accounting.
- A bug bounty incentivizes vulnerability disclosure before exploitation.
- Public reports record discovered bug classes and accepted changes.
- Limits and emergency pause
- Pause mode stops new transfers when an anomaly is detected.
- Release limits and message issuance limits cap the damage from false confirmation.
- Separation of governance roles reduces concentration of authority.
- ZK proofs and light clients
- A ZK approach verifies a cryptographic proof of an event instead of trusting a single message delivery provider.
- Light clients verify another chain’s block headers and confirm state without an external server.
- Implementation complexity and computation cost limit adoption, but they reduce the share of trusted components.
Example: in an optimistic scheme, the confirming participant posts collateral; if a false message is successfully challenged, the collateral is slashed and the release on the destination chain is cancelled.
These measures lower the probability of an exploit, but the base condition remains the same: the reserve must be preserved and cross-chain message confirmation must be correct.
Major incidents show two recurring scenarios: false confirmation and governance compromise.
Largest crypto bridge hacks: typical failure points
In major exploits, tokens were released on the destination chain without a valid transfer on the source chain. This happened when the confirmation mechanism accepted a false signal because of key compromise, a signature verification bug or an incorrect smart-contract state.
- Ronin (March 2022, about $620 million). Compromise of 5 of 9 validators allowed locked assets to be withdrawn; the root risk was governance concentration among a limited set of operators.
- Poly Network (August 2021, about $611 million). A contract logic bug allowed arbitrary cross-chain messages to be formed and assets to be withdrawn from several networks.
- Wormhole (February 2022, about $325 million). A signature verification bug led to the minting of unbacked wETH.
- Nomad (August 2022, about $190 million). A deployment mistake caused the contract to accept any transfer as valid. Users copied one successful transaction and withdrew funds from the contract, causing a mass liquidity drain.
- BSC Token Hub (October 2022, about $570 million). A vulnerability allowed new BNB to be minted through the bridge; limiting the damage required halting the network.
Attacks concentrate on message confirmation and key management, so signature thresholds, release limits and pause mode reduce the maximum damage during a failure.
Bridge security is defined by the resilience of cross-chain message confirmation and by protection of upgrade governance.
Before a transfer, it is important to check the route and understand where the bridge may stop release or lose liquidity.
Checks before a cross-chain transfer
- Test transaction
- A small amount checks the network, token, recipient address and actual receipt on the destination chain.
- A test transfer reveals a wrong network or token choice before moving a large amount.
- Domain and interface-source check
- Phishing sites replace the domain and are promoted through ads and comments.
- Theft often happens through a malicious approval followed by token spending through that permission, or transferFrom.
- Liquidity and release limits
- Liquidity pools require reserves on the destination chain for release.
- Pause mode, release limits and congestion messages reflect transfer limitations at the protocol level.
- Wrapped token custody risk
- A wrapped asset depends on the reserve and bridge governance, so its risk differs from the native asset on the destination chain.
- A bridge halt reduces wrapper liquidity and worsens redemption conditions.
- Split amount and routes
- Splitting the amount limits damage from a single mistake or a release failure.
- Using independent routes reduces concentration of risk on one reserve and one confirmation mechanism.
- Protocol status and pause modes
- Release halt on the destination chain is implemented through a protocol pause or message blocking.
- During incidents, risk rises because limits may change and routes may be disabled.
- Separate storage and operating addresses
- Operating addresses interact with third-party contracts and approvals, so their risk profile is higher.
- A storage address does not interact with third-party smart contracts and does not grant token spending approvals, reducing the number of theft points.
The checks apply to any DeFi operation that uses approvals and interaction with external contracts.
Losses during cross-chain transfers are most often tied to interface substitution, insufficient liquidity on the destination chain and message confirmation failures.
❓ FAQ about crypto bridges
What is a crypto bridge in simple terms?
A crypto bridge moves a token from blockchain A to blockchain B. In lock-and-mint, the token is locked on the source chain and a wrapped equivalent is issued on the destination chain. In the liquidity-pool model, release on the destination chain comes from a pool after the event is confirmed on the source chain.
Why are crypto bridges riskier than transactions inside one network?
A cross-chain transfer adds trusted components: the reserve contract, the message confirmation mechanism and upgrade governance. Compromise of any of these components can lead to unbacked assets being minted or the bridge reserve being withdrawn.
What signs are used to assess the reliability of a crypto bridge?
The assessment comes down to three signs: the reserve contract address, the interchain message confirmation mechanism such as multisig, guardians, PoS validators or oracle plus relayer, and the upgrade governance rules such as signature threshold, timelock and release limits.
What usually happens after a crypto bridge hack or transfer halt?
After an incident, the bridge team stops new transfers or introduces limits to stop losses. Wrapped tokens on another chain often trade at a discount to the original asset because redemption becomes uncertain. Recovery includes fixing the bug, revising governance and attempting to close the reserve deficit.
What alternatives are used for moving assets between networks?
Practical alternatives include withdrawals through centralized exchanges, with custody risk and compliance requirements, and native cross-chain solutions within individual ecosystems, with limits by network and token. Choosing an alternative changes the trusted surface but does not remove risk completely.
What matters about crypto bridges
A crypto bridge moves a token from one network to another without selling the asset. Bridges are used to work with applications on another network and reduce fees on the destination chain.
The key risk of a cross-chain transfer is tied to the reserve contract and the message confirmation mechanism. A verification error, key-threshold compromise or incorrect parameter upgrade leads either to reserve withdrawal or to unbacked wrapped assets being minted.