Crypto Bridges Explained, Including Why So Many Get Hacked

What a bridge does, the three main designs, why they hold the record for the largest hacks in crypto, and how to use one with less risk.

Last reviewed 2026-09-14

The problem bridges solve

Blockchains cannot see each other. Ethereum has no way to know what happened on Arbitrum, and no mechanism to move a token from one to the other. A bridge is the workaround: it locks an asset on the source chain and issues a claim on the destination chain.

Critically, nothing crosses. Your ETH does not travel to Arbitrum. It is locked on Ethereum while a representation of it is minted on Arbitrum. The representation is only worth something because the lock holds.

Three designs, three risk profiles

Native rollup bridges. Arbitrum, Optimism and Base each have an official bridge secured by the rollup’s own proof system. This is the safest category — the bridge is part of the chain’s security model rather than a separate system. The cost is time: withdrawing from an optimistic rollup takes seven days unless you pay a third party to front the funds.

Lock-and-mint bridges. A contract or a multisig holds the deposits; a token is minted on the far side. Security depends entirely on whoever controls the lock. This category has produced nearly every large bridge hack.

Liquidity network bridges. Pools exist on both chains and a relayer pays you out of the destination pool. Nothing is minted, so a failure costs the liquidity providers rather than breaking the token’s backing. Faster, and usually more expensive.

Why the hacks are so large

The four biggest exploits in crypto history are all bridges:

• Ronin, 2022 — $624m. Five of nine validator keys compromised through social engineering.
• Poly Network, 2021 — $611m. A function-call flaw let the attacker reassign the contract’s keeper.
• BNB Bridge, 2022 — $586m. A forged Merkle proof minted two million BNB from nothing.
• Wormhole, 2022 — $326m. A signature-verification bug allowed minting without a deposit.

The pattern is structural. A bridge is a single contract holding the collateral for every wrapped token it has issued, which makes it the highest-value target on the chain, while being newer and less battle-tested than the chains it connects. Vitalik Buterin’s argument that security does not extend across chains — that a 51% attack on a small chain can steal from a bridged asset on a large one — remains the clearest statement of the problem.

Using one with less risk

1. Prefer the official bridge. For any rollup, the chain’s own bridge is almost always the safest route.
2. Check how long it has existed and how much it holds. A bridge that has held $2bn for three years has survived more attack attempts than a new one with better marketing.
3. Do not leave funds mid-route. Bridging is a transition, not a place to store money.
4. Test with a small amount first on any bridge you have not used before.
5. Consider not bridging at all. Withdrawing from an exchange directly onto the destination chain is often cheaper and avoids the bridge contract entirely.

The wrapped-token footnote

When you hold a bridged asset, you hold a claim on a contract. If the bridge is drained, the wrapped token on the destination chain loses its backing and its price goes to roughly zero, while the original asset sits untouched on the source chain in the attacker’s hands. “Wrapped” versions of the same asset from different bridges are different tokens with different risk — check which one you actually hold.