Skip to content
Reading depth

Each view includes the earlier layers; the complete engineer or auditor page is shown by default.

Bridge risk is the possibility that cross-chain messages or asset representations become unauthorized, duplicated, delayed, censored, or no longer backed.

Warning cross category cue

Bridge risk is the possibility that a cross-chain message or destination asset no longer matches authorized and finalized source-chain state.

Separate blockchains do not natively share one state or finality rule. A bridge must define how a source event is observed, finalized, authenticated, relayed, and consumed. Asset bridges must additionally reconcile what is locked, burned, minted, released, or supplied as route liquidity.

Cross-system settlement and correspondent custody are useful analogies. Bridges also combine independent consensus, reorganization, sequencer, verifier, contract, upgrade, and message-ordering assumptions. A forged destination token can immediately become collateral or pool inventory before manual reconciliation.

Source assets are locked, burned, or provided as route inventory. Destination assets are released or minted. The destination token or pending message is a claim whose meaning depends on the bridge’s verification and backing model.

  1. A source transaction locks or burns an asset and emits a message.
  2. The bridge waits for its selected source-finality condition.
  3. A verifier authenticates the chain, sender, destination, payload, and nonce.
  4. A relayer delivers the message to the destination contract.
  5. The destination releases or mints value and marks the message consumed.
  6. A reverse path or liquidity provider later completes redemption where supported.

Information flow carries authenticated messages. Capital flow locks, burns, releases, or routes assets. Claim flow creates the destination representation. Return flow can pay relayers or liquidity providers from user fees or explicit incentives, but it does not prove backing or safe redemption. Risk flow spreads from an invalid message or backing gap into every protocol that accepts the representation.

Bridges may charge users or pay liquidity providers and relayers. Those fees compensate service and capital but do not guarantee backing. Loss can reach the sender, recipient, liquidity provider, wrapped-token holder, downstream lender, or an explicit reserve according to the failure and recovery rules.

Ethereum.org’s bridge documentation compares lock-and-mint, burn-and-mint, and liquidity-network patterns and names their differing verification and custody assumptions. It is a design overview, not an endorsement of any bridge or a guarantee that “trustless” removes all implementation and chain risk. The documentation was reviewed 2026-08-10.

An engineer or auditor should verify domain separation, source and destination identity, sender authentication, payload encoding, nonces, replay protection, finality, reorg handling, validator thresholds, rate limits, supply backing, paused and delayed states, upgrade authority, and cross-chain recovery ordering.

  • “Assets move between chains.” Many designs lock or burn value on one chain and create a separate representation or release on another.
  • “A valid proof proves the asset is backed.” Proof validity, message meaning, custody, supply accounting, and redeemability are separate properties.
  • “Canonical bridge means risk-free.” Canonical names do not eliminate contract, governance, chain, sequencer, or integration failures.

Inspect the underlying wrapped asset and integration risk.

Machine-readable model

Key equations

Canonical expressions come from the structured concept record. KaTeX renders the notation, while the plain-text expression and variable table keep its meaning and units inspectable without JavaScript. Read the narrative above for the model's domain, assumptions, and rounding rules.

This concept does not require one canonical equation. Its mechanism and state transitions remain the authoritative explanation; do not invent a formula merely to make the topic look quantitative.

Assurance contract

Security properties

These structured statements define desired behavior. Their stable IDs can bind tests, invariants, specifications, audit findings, or proof results without turning descriptive review advice into an assurance claim.

Desired · not evaluated: No test, audit, or proof result is implied until scoped evidence is linked to this property.

  1. Each accepted message is authenticated to the intended source chain, sender, destination, payload, nonce, and finality rule and is consumed at most once

  2. Minted or released destination value reconciles with locked, burned, escrowed, or liquidity-provided source value under the stated bridge model

  3. Reorganization, conflicting messages, paused chains, delayed relayers, and unavailable proofs enter explicit pending or recovery states

  4. Validator thresholds, custody, rate limits, guardians, upgrades, and emergency recovery cannot silently exceed their disclosed authority

Knowledge check

Quiz

Answer in your own words, then open the model answer.

What problem does Bridge risk exist to address?

Model answer

Separate blockchains do not share one native state, so a bridge must define which source event is final, who verifies it, and how destination claims reconcile with locked or burned assets.