Why 'Trust-Minimized' Bridges Are Still a Distant Dream

The multisig is the vulnerability. The dominant security model for bridges like Stargate and Multichain is a permissioned multisig. This creates a single point of failure, as evidenced by the $625M Ronin Bridge and $200M Wormhole exploits, which targeted validator keys.

Light clients are not a panacea. Projects like IBC and Near's Rainbow Bridge use light clients for cryptographic verification, but they trade validator risk for liveness assumptions. A chain halt or a 51% attack on the source chain can still compromise funds.

Economic security is misaligned. Models like Across's optimistic verification or LayerZero's Oracle/Relayer design externalize risk. They rely on third-party actors staking capital, creating security that is probabilistic and reactive, not deterministic.

Evidence: Over $2.5 billion has been stolen from cross-chain bridges since 2020, making them the most exploited crypto primitive. No major bridge has achieved the trust-minimized security of its underlying blockchains.

On-chain verification is prohibitively expensive. A light client must verify cryptographic proofs of the source chain's consensus on the destination chain. For proof-of-stake chains like Ethereum, this involves verifying thousands of signatures per block, a gas cost that destroys economic viability for all but the largest value transfers.

The gas cost asymmetry is permanent. The cost to produce a validity proof on a rollup is trivial; the cost to verify it on Ethereum is immense. This asymmetry creates a fundamental scaling limit for protocols like IBC or zkBridge that require on-chain light clients, confining them to niche, high-value corridors.

LayerZero's oracle-relayer model sidesteps this wall. By outsourcing verification to an off-chain network and posting only a hash on-chain, LayerZero and Stargate avoid the gas cost entirely. This trade-off replaces cryptographic certainty with cryptoeconomic security, which is the pragmatic standard for today's bridges.

Evidence: The Succinct Light Client for Ethereum on Gnosis Chain costs ~500k gas to verify a single epoch. Verifying a single Ethereum block header directly on another EVM chain can cost over 3 million gas, making per-transaction bridging impossible.

ZK light clients are latency-bound. A validity proof for a single Ethereum block takes ~20 minutes to generate on consumer hardware, making real-time cross-chain messaging impossible for applications like Uniswap.

Shared sequencers centralize trust. Networks like Espresso or Astria propose a single sequencer set for multiple rollups, but this creates a new, monolithic trust point for bridges like LayerZero that defeats the purpose of decentralization.

Economic security is misaligned. The cost of a zero-knowledge proof for a full block's state dwarfs the value of a typical bridge transaction, making the model economically non-viable for protocols like Across or Stargate.

Evidence: The fastest ZK-EVM, Polygon zkEVM, has a 5-hour finality window for Ethereum; shared sequencer designs explicitly trade off liveness for throughput, a fatal flaw for bridge security.

Universal trustlessness remains impossible without a shared, battle-tested light client or zero-knowledge proof system for every major chain. The technical fragmentation between heterogeneous chains like Ethereum, Solana, and Cosmos creates a coordination nightmare no single protocol like LayerZero or Axelar can solve.

The pragmatic path is niche optimization. Protocols must choose between security, speed, and cost. Across Protocol optimizes for security with bonded relayers and slow, optimistic verification. Stargate optimizes for unified liquidity and speed via the LayerZero Omnichain Fungible Token standard, accepting different trust assumptions.

Evidence: The Total Value Locked (TVL) and attack surface are directly correlated. The 2022 Wormhole and Nomad hacks, which lost over $1 billion, exploited the trusted verification models that enable cross-chain speed and liquidity today. True decentralization sacrifices scalability.

Future-proofing means building for intent. The endgame is user-centric routing, not chain-centric bridges. Systems like UniswapX and CowSwap abstract bridge choice, allowing aggregators like Across to compete on execution. The bridge becomes a commodity, and the intent-based solver network becomes the critical, trust-minimized layer.