The security of that release depends on Decentralized Verifier Networks —off-chain nodes that attest that the message is valid. Kelp DAO configured its bridge with a 1-of-1 DVN threshold, meaning a single verifier was enough to authorize any cross-chain message .
The attacker compromised Kelp’s internal RPC nodes and DDoS’d external nodes, leaving only that single verifier operational and feeding it a forged message that claimed 116,500 rsETH had been burned on the source chain. The verifier attested to the message. The Ethereum contract obeyed. Funds were released to an attacker-controlled address .
Chainalysis confirmed that every on-chain transaction appeared legitimate to standard security tools, because the breach happened entirely off-chain at the infrastructure and node level . Traditional smart contract audits were irrelevant.
Kelp’s emergency multisig paused the contracts 46 minutes after the initial drain, preventing an additional ~$200 million in follow-up attacks .
The attacker didn’t sit on the stolen tokens. Within hours, 89,567 of the 116,500 unbacked rsETH had been deposited into Aave V3 as collateral, and the attacker borrowed roughly 82,650 WETH and 821 wstETH—clean, liquid assets—before anyone could freeze the positions . Similar collateralized borrowing occurred on Compound and Euler, extracting approximately 74,000 clean ETH total
.
Then the laundering began in earnest.
Over the following six weeks, the attacker laundered nearly all of the unfrozen stolen funds—approximately $220 million—leaving only about $1.7 million traceable in the original exploiter wallets as of June 1, 2026 . The laundering chain followed a deliberate two-stage pattern:
TRM Labs later confirmed that THORChain operated as the consistent bridge of choice across North Korea’s largest heists, with no operator willing to freeze or reject transfers during either the 2025 Bybit breach or the KelpDAO exploit .
NS3.AI also flagged a novel detail: the attackers used LayerZero itself to move at least $500,000 of the stolen funds across chains during the laundering phase—marking the first recorded instance where the same application was exploited for both the theft and part of the money laundering .
Not all the funds escaped. On April 20, 2026, at 11:26 PM ET, the Arbitrum Security Council executed an emergency action to freeze 30,766 ETH—approximately $71 million, or roughly a quarter of the total stolen amount—held in an attacker-controlled address on Arbitrum One .
The Council acted with input from law enforcement and moved the funds to a governance-controlled intermediary wallet. Nine of 12 council members voted in favor of the freeze . The funds can only be released through a formal Arbitrum governance vote
.
On May 8, 2026, the Arbitrum Security Council approved a joint proposal to unfreeze those funds, aiming to accelerate rsETH collateral recovery and restore liquidity for affected users. The recovery process remains ongoing with law enforcement involvement .
Aave absorbed the most severe second-order damage. The attacker deposited 89,567 fake rsETH into Aave V3 and borrowed approximately $230 million in clean assets—loans that became unrecoverable bad debt once the rsETH was revealed as unbacked .
Aave’s Protocol Guardian froze rsETH and wrsETH reserves across all V3 deployments at approximately 19:00 UTC on April 18, setting loan-to-value ratios to zero across 11 affected markets including Ethereum, Arbitrum, Avalanche, and Optimism . WETH borrowing—a core piece of DeFi’s financial plumbing—was effectively frozen across six networks.
As of mid-May 2026, over 95% of the unbacked tokens had been recovered, with the remaining shortfall expected to be covered by the Aave DAO treasury and the DeFi United coalition . Aave restored normal wETH borrowing limits across six V3 networks on May 18, 2026
.
But the real legacy is the governance response. At Consensus Miami 2026, Aave Labs Chief Legal and Policy Officer Linda Jeng announced a fundamental overhaul of the protocol’s asset listing and collateral evaluation standards . The new framework expands beyond traditional financial risk metrics to include:
Aave has already adjusted 295 risk parameters and added automated defenses that can reduce an asset’s loan-to-value ratio to zero when predefined risk thresholds are triggered . The protocol is launching a full review of every asset listed on V3 and rewriting its listing standards from the ground up
.
Kelp DAO didn’t happen in isolation. It was the second nine-figure bridge exploit in 18 days, following the Drift Protocol’s $285 million social-engineered breach on April 1—also attributed to Lazarus Group . Combined, those two incidents pushed early-2026 DeFi losses past $840 million
.
The systemic fallout dwarfed the direct theft. Within 48 hours of the Kelp DAO exploit, $13.21 billion in total value locked evaporated across DeFi, with Aave alone losing 43% of its TVL across 26 tracked protocols . A $5.4 billion withdrawal panic swept the ecosystem
.
The attack exposed what Chainalysis called a critical structural blind spot: DeFi security had focused overwhelmingly on smart contract audits while bridge infrastructure, node operational security, and single-verifier configurations remained largely unexamined risk vectors .
The fix is already underway. Protocols are migrating to multi-verifier bridge configurations. Aave’s new listing handbook—expected to be published as a formal playbook for asset issuers—will require projects to disclose bridge architecture, verifier decentralization, and node security practices before rsETH-like derivatives can be onboarded as collateral .
Lazarus exploited a gap between what DeFi audited and what DeFi actually depended on. The industry’s response suggests that gap is finally closing—but only after a $293 million lesson.
Comments
0 comments