Multisig validators sign separately, while MPC validators use key shares to produce one signature. Both can require a threshold before a bridge releases a transfer, but they expose different details to people checking the chain.
How bridge validator signing works
With multisig, each validator signs the transfer with its own key. The destination contract checks the signatures against the current validator set and releases the funds once enough valid signatures meet the threshold. The transaction can show which validators signed.
With MPC, validators hold shares of one private key. They work together to create a single ordinary signature; no participant holds the complete key. The destination contract verifies that signature as it would a regular signature, so it does not see the individual participants.
Say a bridge must approve a transfer from one network to another. In a multisig setup, the destination contract receives separate validator signatures and checks whether the required threshold has signed. In an MPC setup, the validators combine their key shares to produce one signature for that same transfer. The destination still checks authorization before releasing funds; what changes is the signing process and what the chain reveals about it.
That visibility matters when you assess a bridge. Multisig signatures and validator-set changes can be inspected on-chain, making the signers and threshold easier to audit publicly. MPC usually costs less gas because the destination verifies one signature, and it hides the signer count. The chain does not reveal the MPC participants, threshold, or key resharing; checking those details depends on off-chain disclosure.
cBridge uses a different, visible verification model: its destination contracts check individual State Guardian Network (SGN) validator signatures against a stake-weighted quorum. For background on how cBridge compares with other bridge designs, cbridge alternatives adds useful context on the options around this choice.
The Ronin Network bridge exploit post-mortem (2022) reported that attackers compromised five of the bridge’s nine validator keys. That case shows why the number and control of signing keys matter: a threshold can be reached if enough keys are compromised. It does not make the signing model alone a complete measure of bridge security.
Gennaro and Goldfeder’s 2018 study, Fast Multiparty Threshold ECDSA, describes threshold ECDSA protocols that produce one standard signature from key shares. Its finding concerns the cryptographic signing protocol; it does not establish how a particular bridge manages validators or protects transfers.
Related
What happens when a bridge's signing keys are stolen?
Stolen keys let an attacker sign only if they control enough authority to meet the bridge’s threshold. With multisig, signatures identify the keys used on-chain. With MPC, the chain shows one signature, so identifying compromised participants depends on information disclosed off-chain.
Can a bridge change its validator set?
Yes. A bridge can update which validators are authorized, subject to its governance and contract rules. Multisig set changes can be visible on-chain; MPC participant changes and resharing may not be. When choosing between them, look for how the bridge discloses its signers, threshold, and changes over time.
Choose the model whose signer information you can verify to the level your risk decision requires.