Is cBridge Decentralized or Custodial? Who Controls Funds
cBridge is non-custodial but not trustless: users send assets into smart contracts, while a quorum of State Guardian Network validators authorizes releases on the destination chain. The common belief that “decentralized” means no outside party must be trusted misses that release decision. The cBridge site provides context for the Celer bridge as a way to move supported assets between networks; the custody question begins with who can release them.
No company account receives the user’s deposit. That is the practical meaning of non-custodial here. But funds being held by a contract does not tell you who can instruct that contract to pay out. To assess the Celer bridge, follow both the assets and the authority to release them. Those paths lead to different answers.
How cBridge Authorizes a Release
cBridge uses a source-chain transaction, validator signatures, and a destination-chain contract to complete a transfer. The user authorizes the source transaction from their own wallet. State Guardian Network validators observe it, verify the cross-chain event, and co-sign the destination release. The destination contract checks that the signatures represent more than two-thirds of the validators’ staked power before releasing pool funds or minting a pegged token.
- User: Signs the transaction that sends assets to a bridge contract or liquidity pool. The user controls that wallet action, rather than handing assets to a company account.
- Bridge contracts: Hold locked assets or manage liquidity pools. Their rules determine what a valid destination release can do with those assets.
- State Guardian Network validators: Stake CELR and sign messages about source-chain transactions. Their stake-weighted quorum supplies the external approval that the destination contract accepts.
- Governance: Holds controls that can pause activity, set delays for large transfers, and cap volume. These controls affect when and how much the contracts can process.
The detail to watch is who signs a release. Picture two bridges with the same wallet flow and the same smart contracts. In one, a single company key can authorize a destination payout. In the other, approval requires signatures from validators holding more than two-thirds of staked power. Changing only that signing rule changes who has release authority: one company in the first case, a distributed but externally trusted validator set in the second. cBridge follows the second model.
That distinction also explains the risk. If enough validator keys were compromised, or their operators colluded, they could sign an improper release of pooled or locked funds. The contract would see the required quorum, even though the underlying transfer did not justify the payout. Governance controls such as pauses, delays, and volume caps can constrain activity, but they are separate from the signature rule and bring their own control over the bridge.
Compare this with a canonical rollup bridge to Ethereum. A canonical bridge relies on the rollup’s own proof system to establish what happened before assets are released. The Celer bridge instead asks an external validator set to attest to the source event. Both may use contracts, yet the evidence those contracts accept differs. Treating “held in a contract” as equivalent to “verified by the chain’s own proofs” would hide the extra trust placed in validators.
Nor does “non-custodial” mean a transfer has no intermediate exposure. During an xAsset transfer, original tokens remain locked on the source chain, backing a pegged token minted on the destination. During an xLiquidity transfer, liquidity pools supply the asset released on the destination. In either case, the user’s wallet starts the transfer, while the destination outcome depends on the validator quorum and the relevant contract rules.
Related Questions
What is the State Guardian Network?
The State Guardian Network, or SGN, is the group of CELR-staking validators that verifies source-chain events and signs cross-chain release messages. Its signatures are weighted by stake, so the decisive threshold is more than two-thirds of staked power, not simply a count of individual validators. This is why learning that “multiple validators sign” is only a start: the share of stake behind those signatures is what the destination contract checks.
Do you need an account to use cBridge?
cBridge’s non-custodial model centers on transactions signed from the user’s own wallet rather than a deposit into a company account. An account question is therefore different from the release-authority question. Having control of the wallet that starts a transfer tells you who authorizes the send; the SGN signature requirement tells you who authorizes the destination payout. Both matter when asking who controls funds in transit.
What is the difference between xAsset and xLiquidity?
xAsset uses a lock-and-mint arrangement: original tokens stay locked on the source chain, and a pegged token is minted on the destination against that lock. xLiquidity uses liquidity pools: the destination release draws on assets already available in a pool. The observable result differs—a pegged token in one model, pool liquidity in the other—but both rely on SGN validators to approve the cross-chain release. The asset route changes; the external trust assumption remains.
cBridge sits between exchange custody and a bridge that relies on a chain’s own proofs. Users keep control of their wallets and send assets to contracts, while SGN validators with more than two-thirds of staked power must sign destination releases. That signing rule is the deciding reason to call the bridge non-custodial and externally trusted.