Bridge risk
how cross-chain infrastructure failures drain protocols
Cross-chain bridges are the highest-loss attack surface in DeFi. Multisig validators, lock-and-mint pools, and upgrade-key concentration each create distinct vulnerabilities. Understanding the trust model of every bridge in your dependency chain is a prerequisite for assessing protocol risk.
Analyse a protocolBridge dependency analysis · upgrade authority · EVM
Every scan is powered by the XemaS Semantic Intelligence Platform · answers carry their evidence and coverage state
Bridges concentrate assets and trust in a single point
A cross-chain bridge must lock or pool assets on one side and mint or release them on the other. This creates a large, static pool of assets guarded by whatever mechanism the bridge uses to validate cross-chain messages. That mechanism - whether a multisig, a validator set, or a cryptographic proof system - is the bridge's security model.
When the security model fails, an attacker can forge valid cross-chain messages and unlock the entire locked pool without depositing anything on the source chain. The assets have effectively been minted without backing. Every user who holds the bridged token is exposed to a loss of peg.
Bridge risk is also a dependency risk. Any protocol that accepts bridged assets as collateral, or that routes cross-chain through a specific bridge, has embedded bridge risk in its operations regardless of how secure its own contracts are.
Five bridge designs and their risk profiles
| Bridge type | How it works | Risk level | Primary concern |
|---|---|---|---|
| Lock-and-mint bridge | Locks native assets on source chain, mints wrapped tokens on destination | High | Locked asset pool is a single high-value target; mint logic must be identical to lock logic |
| Liquidity network bridge | Transfers native assets from destination-chain liquidity pools | Medium | Pool imbalances create withdrawal delays; LP exposure to bridge hacks |
| Optimistic bridge | Assumes transactions are valid; fraud proofs can challenge within a window | Medium | Challenge window (typically 7 days) creates exit delays; fraud proofs require active watchers |
| ZK proof bridge | Generates zero-knowledge proofs of source-chain state for destination verification | Low | Circuit bugs can invalidate the security model; prover infrastructure is centralised in most deployments |
| Multisig validator bridge | A threshold set of validators attest to cross-chain events | High | Security collapses if the signing threshold is compromised; the majority of large bridge hacks used this model |
Five-step bridge risk assessment
Identify the bridge type and trust model
Is the bridge secured by a multisig, a set of validators, an optimistic fraud-proof window, or a zero-knowledge proof? Each model has a different trust assumption. The question is: what does an attacker need to compromise to drain the bridge?
Check the validator or signer threshold
For multisig and validator bridges, identify the signing threshold and the total number of signers. A 5-of-9 multisig with nine publicly disclosed independent validators is materially different from a 3-of-5 with three team members. Verify both the threshold and the identity independence of signers.
Check the bridge contract upgrade authority
Can the bridge contract be upgraded? If so, who controls the upgrade key? A bridge with $500M in locked assets whose upgrade key is held by a single EOA is one private key loss away from a complete drain. Verify whether upgrades require governance approval and whether a timelock is in place.
Assess how much the protocol depends on this bridge
What percentage of the protocol's TVL or functionality flows through the bridge? A protocol that requires a specific bridge for its core operations has bridge risk embedded in its operational model. Identify the dependency and whether alternative paths exist.
Verify whether the bridge has been audited recently
Bridge security is complex and multi-chain. An audit from two years ago on a bridge that has since added new chains or upgraded its contracts provides limited assurance. Check whether audit coverage is current and whether it covers the specific chains and contract versions in use.
Free · No sign-up required
What analysts get wrong about bridge risk
Assuming a bridge is safe because it is large
The largest bridge exploits have involved bridges with hundreds of millions in TVL. Scale creates incentive for attackers, not resistance to them. Bridge TVL measures how much is at risk.
Not distinguishing wrapped assets from native assets
An asset bridged via a lock-and-mint bridge is a wrapped representation of the original. If the bridge is exploited and the locked assets are drained, the wrapped tokens on the destination chain may become worthless. Native assets held directly on-chain carry no bridge counterparty risk.
Treating validator count as validator independence
A bridge secured by 9 validators where all 9 use the same infrastructure provider and all are employees of the same company has weaker practical security than its threshold implies. Independence of signers matters as much as the count.
Ignoring bridge risk in composable protocols
A DeFi protocol that sources liquidity from another chain, or accepts bridged assets as collateral, inherits the bridge risk of the assets it accepts. The bridge risk is upstream in the dependency chain, not visible in the protocol's own contracts.