Custody Models Behind Bitcoin Bridge Risk

A single entity holds all bridged assets, creating a single point of failure and censorship risk. This model underpins most major CEX-operated bridges and early custodial services.

• Attack Surface: Compromise of a single private key or legal seizure.
• Trust Assumption: Users must trust the custodian's security and solvency.
• Examples: Wrapped Bitcoin (WBTC) via BitGo, early iterations of RenVM.

Key control is distributed among a permissioned set of nodes, reducing reliance on a single actor. This is the dominant model for modern institutional bridges.

• Security Model: Requires a threshold (e.g., t-of-n) of nodes to collude for theft.
• Trade-off: Introduces validator risk and potential for governance capture.
• Examples: Threshold Network (tBTC v2), Multichain (prior to collapse), Liquality.

Security is derived directly from Bitcoin's consensus, eliminating intermediary validators. This is the holy grail but faces significant technical hurdles.

• Mechanism: Uses SPV proofs or light clients to verify Bitcoin state on the destination chain.
• Limitation: High gas costs and complex integration with smart contract chains.
• Examples: Babylon (staking), Bitcoin Spark, Nomic (experimental).

Accepts a weaker custody model but backstops it with robust economic security. This aligns incentives for federations and provides user recourse.

• Primary Backstop: Overcollateralization by node operators (e.g., 150-200%).
• Secondary Layer: Protocol-owned insurance funds or crowdsourced coverage.
• Examples: tBTC v2 (operator bonds), Across (optimistic validation + liquidity pool).

A council of known entities controls a multisig wallet on the destination chain. This model, used by Wrapped Bitcoin (WBTC) and early bridges, creates a single point of political and technical failure.\n- Risk: $10B+ TVL contingent on the honesty and operational security of a few entities.\n- Failure Mode: Collusion, regulatory seizure, or a single signer's private key compromise can drain the entire reserve.

Attempts to create a trust-minimized bridge by verifying Bitcoin block headers on the destination chain. Projects like Babylon and tBTC v2 explore this. The model shifts risk from validators to cryptographic assumptions and economic incentives.\n- Risk: Relies on a live, well-funded watchtower network to submit fraud proofs.\n- Failure Mode: Insufficient staking, protocol bugs in the light client, or long-range attacks on Bitcoin's consensus can invalidate proofs.

Avoids direct custody by using hashed timelock contracts (HTLCs) and liquidity pools, as seen in Thorchain and Liquid Network peg-in/out. Users trade native BTC for a synthetic asset via a decentralized vault system.\n- Risk: Concentrated in the bonded node operators who custody the vaults and the stability of the underlying Proof-of-Stake chain.\n- Failure Mode: A >33% Byzantine node cohort can halt or censor transactions; complex cross-chain arbitrage can drain liquidity pools.

Uses Multi-Party Computation (MPC) or Threshold Signature Schemes (TSS) to distribute key generation and signing across a decentralized network, as implemented by Keep Network (tBTC v1). Reduces single points of failure but introduces new complexities.\n- Risk: Security depends on the correct implementation of complex cryptographic protocols and the randomness used in key generation.\n- Failure Mode: A flaw in the MPC library, a compromised random beacon, or a malicious majority of signers can lead to total loss of funds.

A council of known entities (e.g., WBTC, Multichain) holds the BTC. This is the dominant model with ~$10B+ TVL but is a systemic risk vector.\n- Single Point of Failure: Compromise of >1/3 of signers can drain the vault.\n- Regulatory Attack Surface: Centralized entities are KYC/AML targets, enabling censorship.\n- Opaque Governance: User trust is placed in off-chain legal agreements, not cryptographic proofs.

Projects like Babylon and Nomic use Bitcoin's consensus directly. They verify SPV proofs or zk-SNARKs of Bitcoin state on the destination chain.\n- Trust Minimized: Security reduces to Bitcoin's 51% attack cost, not a new federation.\n- Sovereign Verification: Each chain validates the proof; no new social consensus needed.\n- High Latency Cost: Native Bitcoin finality is ~1 hour, making this model slow for high-frequency swaps.

Protocols like Threshold Network (tBTC) and Interlay use overcollateralized, dynamically selected signer sets. It's a fusion of crypto-economic and federated models.\n- Bonded Custodians: Signers must stake the native token (KEEP/INTR), slashed for malice.\n- Decentralized Selection: Operators are permissionlessly chosen from a pool, reducing cabal risk.\n- Complexity Trade-off: Introduces oracle risk for BTC price feeds and liveness dependencies.

WBTC dominates because liquidity begets liquidity. Its federated risk is accepted for DeFi composability on Ethereum.\n- Network Effect Lock-in: Major protocols (MakerDAO, Aave) integrate the deepest pool, creating a moat.\n- Architect's Dilemma: Building a new DeFi primitive? You must integrate WBTC, perpetuating its dominance.\n- The Escape Hatch: Layer 2s and Alt-L1s are the greenfield for native, trust-minimized bridges like Babylon.