KETJU Research

← The Register

other

Avalanche Bridge

Rejected
Max sleeve
Reviewed
2026-08-17 · v1
Next review
2026-11-17
Research basis
Individual research
Chains
Ethereum · sovereign

Rejected venues wait the longest for re-review; a rejection has to earn another look before the scheduled date.

REJECTED ON A FOUR-PARTY COMMITTEE SMALLER THAN EVERY OTHER MULTI-PARTY BRIDGE THIS REGISTRY HAS REVIEWED. The Avalanche Bridge, Avalanche’s original Ethereum-to-C-Chain canonical bridge accessed through the Core wallet, secures transfers with only four ”Wardens” using an Intel SGX secure-enclave design; three of the four must report identical transaction data before the enclave mints or releases funds. That approval set is smaller than the already-rejected Gnosis Chain xDAI Stake Bridge’s 4-of-7 validators, and one of the four Wardens is Ava Labs itself, the same organization that built and operates the underlying network — a related-party concentration point none of this registry’s other reviewed bridges carry in the same form. It also adds a distinct hardware-trust dependency, Intel SGX attestation, on top of Warden honesty.

The research file

Mechanism

This is Avalanche’s original Ethereum-to-C-Chain bridge, distinct from Avalanche’s newer Interchain Messaging primitive built for its own subnet ecosystem. Four Wardens — each running an indexing server plus AvalancheGo and Geth nodes — monitor both chains and relay transaction data into an Intel SGX secure enclave. The enclave’s private key is split via Shamir Secret Sharing across the four Wardens, so no single Warden holds it outright, but consensus among only three of the four is sufficient to act.

Control and governance

The four Wardens are Avascan, BwareLabs, Halborn, and Ava Labs itself — meaning one of the four parties securing the bridge is the same organization that built and operates the underlying network, a related-party concentration point this review flags directly. Approval requires three of the four Wardens, a 75% threshold of a four-member committee, reporting identical transaction data before the enclave mints or releases funds. That is a smaller total committee than the already-rejected Gnosis Chain xDAI Stake Bridge’s 4-of-7 validator set, and it layers a second, independent trust assumption on top of Warden honesty: the integrity of Intel SGX hardware attestation, a technology with a documented industry history of side-channel vulnerabilities, not confirmed exploited against this specific bridge but a standing structural dependency regardless.

Incident record

No confirmed exploit of this specific Avalanche Bridge was identified, including against a targeted check of the major 2026 bridge-hack wave already covered elsewhere in this registry, such as the Kelp DAO and Drift Protocol exploits, neither of which named this bridge. Absence of an incident is a bounded negative finding, not proof of durability, given the committee’s small size.

Exit

Transfers take roughly 10 to 15 minutes, dominated by Ethereum finality rather than any Avalanche-side delay — fast under normal conditions, but with no fraud-proof or validity-proof backstop the way rollup bridges have. Correctness rests entirely on three of four Wardens and SGX enclave attestation holding, not on an independent cryptographic or economic guarantee.

Comparison

Against Base, Arbitrum, and Optimism, all approved with limits in this registry and all inheriting Ethereum-native fraud-proof or validity-proof security, this bridge has no rollup-style proof system at all. Against the already-rejected Gnosis Chain xDAI Stake Bridge, this bridge’s 3-of-4 committee is smaller still, and its SGX hardware-trust layer is a distinct risk Gnosis’s bridge does not carry. One of the four approving parties being the network’s own operator is a related-party fact neither Gnosis’s nor the already-rejected Wormhole’s validator sets share in the same way.

Sources

The claims above trace to these. Where a number could not be independently verified, the thesis says so.

Inherited controls

The verdict above grades the protocol layer. Every position also inherits the asset it holds and the chain it settles on. The least safe layer sets the position’s grade, and the position table names which one that is.

ChainVerdictGradeControl constraint
EthereumApproved sovereign No sequencer, no upgrade key, no operator who can be compelled — rule changes require social consensus.
The memo is public. The watching is the product: the terminal reads your clients’ wallets against this Register and flags the events above when they fire. $49 per advisor per month, first 14 days free. Start the trial.