AFX Trade disclosed a roughly $24.15 million loss after an attacker obtained the signing keys used by its bridge validators. The protocol’s on-chain logic accepted the withdrawal because the required quorum of validator signatures was present, so the transaction looked valid to the contract even though the keys behind those signatures had been compromised. This is a classic key-compromise failure mode rather than a break of the bridge contract itself. When authorization depends on offchain validator keys, the security of the entire system hinges on how those keys are stored, accessed, and monitored. If an attacker can sign as a legitimate validator, the contract will often have no way to distinguish malicious authorization from normal operation. Defenders should treat validator keys like production-grade cryptographic root credentials: keep them in hardware-backed systems, split control across multiple independent operators, and rotate them quickly when compromise is suspected. Bridges should also impose withdrawal limits, time delays, and secondary review for unusually large or unusual transfers. Operational monitoring matters as much as cryptography here. Alerts for quorum-level approvals, destination anomalies, and rapid cross-chain movement can slow or stop theft before funds are fully laundered. Incident response plans should assume that signed messages may still be malicious if the signing environment has been breached.