Kelp DAO’s $292M rsETH Bridge Exploit Triggers Aave Freeze and Peg Fears

A $292 million exploit of Kelp DAO’s LayerZero-powered bridge turned rsETH from a yield wrapper into an immediate collateral-risk event across DeFi. With 116,500 rsETH drained from the reserve that backs wrapped deployments on more than 20 networks, traders are now pricing peg risk, liquidity gates, and potential bad debt.

Key Takeaways

• 116,500 rsETH worth roughly $292 million was drained from Kelp DAO’s LayerZero-powered bridge at 17:35 UTC on Saturday.
• The stolen amount equals about 18% of rsETH’s ~630,000 circulating supply tracked by CoinGecko, a supply shock large enough to destabilize redemptions.
• Kelp froze core contracts via its emergency pauser multisig at 18:21 UTC. Two later attempts to pull another 40,000 rsETH each reverted.
• Aave froze rsETH markets on V3 and V4 within hours, and AAVE fell about 10% as traders priced potential bad-debt exposure tied to rsETH.

$292M rsETH Bridge Drain Turns a Yield Token Into a Collateral Event

The cleanest way to frame this exploit is not “Kelp got hacked.” It is “a bridge reserve that collateralized rsETH across more than 20 networks got drained.” That distinction is why the response from lending venues was immediate.

At 17:35 UTC on Saturday, an attacker drained 116,500 rsETH from Kelp DAO’s LayerZero-powered bridge to an attacker-controlled address. The haul was worth roughly $292 million at the time and represented about 18% of rsETH’s ~630,000 circulating supply tracked by CoinGecko.

Kelp’s product design matters here. Kelp DAO is a liquid restaking protocol that routes user-deposited ETH through EigenLayer to earn additional yield and issues rsETH as a tradeable receipt token. Traders use that receipt token as a yield-bearing asset, but also as collateral. Once a token is widely accepted as collateral, the market stops caring about the marketing and starts caring about redemption mechanics and liquidation paths.

What stands out is the reserve location. The bridge that was drained held the rsETH reserve backing wrapped versions of rsETH deployed on more than 20 other blockchains. When that reserve is impaired, the risk is not confined to one venue or one chain. It becomes a question of whether non-Ethereum representations of rsETH remain backed, and whether any unwind forces pressure back onto Ethereum liquidity.

How the LayerZero Packet Led to 116,500 rsETH Leaving the Reserve Bridge

The exploit path, as described by the protocol’s on-chain incident timeline, ran through cross-chain messaging rather than a direct compromise of rsETH itself.

LayerZero is the messaging infrastructure that lets different blockchains send verified instructions to each other. In this case, the attacker “tricked” LayerZero’s cross-chain messaging layer into believing a valid instruction had arrived from another network. That seemingly valid instruction triggered Kelp’s bridge to release 116,500 rsETH to the attacker.

Kelp’s emergency response came 46 minutes later. At 18:21 UTC, the protocol froze core contracts via its emergency pauser multisig. The timing matters because it gives traders a hard boundary for what was and was not possible after the pause.

Two follow-up attempts at 18:26 UTC and 18:28 UTC both reverted. Each carried the same LayerZero packet and attempted to drain another 40,000 rsETH, roughly $100 million per attempt. That sequence supports a practical conclusion for desk risk: the pause likely shut the same exploit path, even if it did not answer the bigger question of how validation was bypassed in the first place.

Kelp publicly acknowledged the incident at 20:10 UTC in its first X post, saying it was investigating with LayerZero, Unichain, its auditors, and outside security specialists. It did not disclose how the exploit bypassed the bridge’s validation logic.

Why Wrapped rsETH Across 20+ Networks Suddenly Faces Backing Questions

rsETH is deployed across more than 20 networks, including Base, Arbitrum, Linea, Blast, Mantle, and Scroll. Cross-chain movement uses LayerZero’s OFT standard.

The key structural point is that the drained bridge reserve was described as backing wrapped versions of rsETH on those non-Ethereum deployments. When the reserve is removed, the market has to reprice every wrapped representation as a claim on something that may no longer be there.

That is where second-order effects show up. If holders on L2s and other networks rush to redeem into ETH on Ethereum, the pressure concentrates on the “unaffected” Ethereum supply and the protocol’s ability to meet withdrawals. The primary risk is not theoretical. It is mechanical: redemption demand can force Kelp to unwind restaking positions to honor withdrawals.

The size of the theft relative to supply amplifies the disorder risk. Losing about 18% of circulating supply is not just a headline number. It is large enough to fragment liquidity across chains and venues, which is exactly how pegs break in practice. You do not need every holder to panic. You need enough cross-chain float to test the redemption pipe at the same time liquidity gates are being installed.

Signals That Decide the Peg: Redemptions, Recovery, and Reopenings

Near-term, the market is trading on operational signals, not narratives.

First, the root-cause disclosure. Kelp has not explained how the bridge’s validation logic was bypassed. Any update that clarifies whether the fix requires contract upgrades or changes to LayerZero message verification will shape how quickly risk committees allow rsETH back into collateral sets.

Second, reopenings and conditions. Aave froze rsETH markets on V3 and V4 within hours. SparkLend froze its rsETH markets, and Fluid froze its rsETH markets as well. The timeline and criteria to reopen matter more than the initial freeze, because reopenings determine whether forced deleveraging ends or extends.

Third, stolen-funds behavior. On-chain movement of the stolen 116,500 rsETH will be watched for signs of obfuscation and any recovery announcements tied to the ongoing investigation with LayerZero, Unichain, auditors, and outside security specialists.

Fourth, broader LayerZero/OFT risk sensitivity. Ethena paused its LayerZero OFT bridges from Ethereum mainnet for roughly six hours as a precaution, stating it has no rsETH exposure and remains more than 101% overcollateralized. That kind of defensive pause is a tell that protocols are treating OFT bridge risk as systemic enough to justify temporary disruption even without direct exposure.

The Trade Is Now About Liquidity Gates and Bad-Debt Math, Not Yield

I’m treating this as a collateral event first and a protocol exploit second. The market’s first reaction supports that read. Aave froze rsETH markets on V3 and V4 within hours, and Stani Kulechov said the exploit was external and Aave’s contracts were not compromised. Even with that reassurance, AAVE still fell about 10% as traders priced potential bad debt tied to rsETH exposure. That is the desk telling you what it cares about.

Scenario one is containment via hard gates. If freezes persist across Aave, SparkLend, and Fluid until Kelp and its partners publish a credible root-cause and remediation path, the immediate bad-debt tail can be capped, but liquidity stays fragmented. Confirmation points are straightforward: explicit reopening criteria, staged parameter changes, and a clear explanation of how the LayerZero validation path was bypassed.

Scenario two is a redemption-driven peg test. The packet’s own framing is blunt: “Whether rsETH holds peg through the weekend depends on how much of the cross-chain float tries to redeem into ETH on Ethereum and whether Kelp can recover any portion of the stolen funds before the Tornado Cash trail goes cold.” If redemptions accelerate while wrapped representations remain in question, the peg becomes a function of available exit liquidity and operational throttles, not the headline APR.

Scenario three is broader OFT de-risking. Ethena’s roughly six-hour pause, despite claiming no rsETH exposure and more than 101% overcollateralization, is the kind of precaution that can spread. If more protocols pause or change parameters around LayerZero OFT bridges, the second-order effect is reduced cross-chain mobility at the exact moment markets want optionality. That tends to widen basis between representations and makes any peg defense harder.

The invalidation point for the “systemic collateral event” thesis is a fast, technically credible root-cause disclosure paired with coordinated reopenings that restore normal redemption and collateral flows without further precautionary pauses. The confirmation point is the opposite: prolonged freezes, continued OFT bridge caution, and visible stress around redemption capacity, because that is when rsETH stops trading like a yield token and starts trading like a liquidity gate with a balance-sheet hole.

Sources

• CoinDesk