Mantle Proposes $69M ETH Loan to Aave, Would Gain Major Governance Stake in Return
Mantle's Core Contributor Team published an early-stage governance proposal on April 24 to lend up to 30,000 ETH to Aave DAO, offering a lifeline to the lending protocol after a $292 million exploit of Kelp DAO's bridge left Aave facing bad debt that, depending on how losses are distributed, could range from approximately $123.7 million to $230.1 million.
Mantle's Core Contributor Team published an early-stage governance proposal on April 24 to lend up to 30,000 ETH to Aave DAO, offering a lifeline to the lending protocol after a $292 million exploit of Kelp DAO's bridge left Aave facing bad debt that, depending on how losses are distributed, could range from approximately $123.7 million to $230.1 million. Mantle itself faces a potential $77.7 million shortfall on its own Aave deployment, which is the clearest explanation for why its team is making the offer. The proposal comes with a notable condition: Aave would hand Mantle delegated voting power over roughly 130,000 AAVE tokens.
The proposal, filed as MIP-34 by Mantle's Core Contributor Team, is still in its discussion phase. It must first clear a Mantle forum poll and a Snapshot vote before Aave DAO considers it separately. If approved on both sides, Mantle's treasury would lend the ETH at an interest rate pegged to approximately Lido's staking yield plus one percentage point, with a repayment window of up to 36 months. Aave DAO would back the arrangement with 5% of its protocol revenue and at least $11 million in AAVE tokens held in a multisig wallet.
What triggered the crisis
The chain of events began on April 18, when attackers exploited a critical weakness in the cross-chain bridge used by Kelp DAO, the issuer of rsETH. Kelp's rsETH is a liquid restaking token, a type of asset that lets users earn Ethereum staking rewards while keeping their funds accessible across DeFi applications. The attacker did not break a smart contract. Instead, they compromised two RPC nodes (the servers that relay blockchain data) and disrupted a third, tricking LayerZero's verification system into approving a fraudulent cross-chain message. The root problem was that Kelp had configured its bridge on the LayerZero network with only a single verifier, meaning one compromised node was enough to forge a valid transaction. LayerZero had previously published integration guidelines recommending a multi-verifier setup.
The exploit released 116,500 rsETH tokens worth roughly $292 million. The attacker then deposited 89,567 of those tokens as collateral across Aave V3 on Ethereum and Arbitrum and borrowed approximately 82,650 WETH (about $190.9 million) and 821 wstETH (about $2.33 million) against them. Kelp froze its contracts 46 minutes after the breach, and Arbitrum's Security Council separately froze around $71 million in related assets. The bulk of stolen funds were routed through THORChain and converted to Bitcoin, making recovery unlikely. Chainalysis and LayerZero both preliminarily attributed the attack to North Korea's Lazarus Group in early findings, though that characterization has not been finalized.
Why Mantle's exposure makes this proposal less than purely charitable
Aave froze rsETH across 11 of its deployments within 90 minutes of the attack. The suspension of WETH borrowing across the Core, Prime, Arbitrum, Base, Mantle, and Linea deployments followed as a separate step with no confirmed timeframe. All affected chains, including Mantle's own Aave V3 deployment, hit 100% WETH utilization, meaning users could not withdraw. Under one of the two bad debt scenarios modeled in Aave's incident report, the Mantle deployment faces a $77.7 million shortfall, equal to 71.45% of all WETH deposits on that chain. That figure, more than any other, explains why Mantle's team framed the loan as converting "idle capital into a yield-generating position while supporting recovery efforts." It also explains the proposal's timing. Mantle had confirmed that its chain, bridges, and core assets were unaffected, but its users on Aave's Mantle deployment are directly exposed.
Mantle's treasury holds more than $4 billion in community-owned assets, which means the 30,000 ETH commitment of roughly $69.4 million, while substantial, represents a modest fraction of available resources. That context matters when weighing whether the proposal is as generous as its framing suggests.
The incident also highlights a risk that other protocols managed to sidestep. SparkLend, the lending protocol associated with MakerDAO, proactively exited the rsETH market three months before the exploit and avoided all losses entirely, demonstrating that the vulnerability was foreseeable through careful, proactive collateral auditing.
The governance angle
Beyond the financial terms, MIP-34 would deliver Mantle delegated voting power over approximately 130,000 AAVE tokens, a significant governance stake in one of DeFi's largest and most widely used lending protocols. The arrangement sets a precedent: a well-capitalized Layer 2 network using a liquidity crisis to secure a meaningful governance position inside a foundational DeFi protocol. Mantle is institutionally linked to Bybit, one of the most widely used crypto exchanges across India, Pakistan, Bangladesh, and Sri Lanka.
What this means for users outside the US
For users across India, Pakistan, Nigeria, and Kenya, the freeze on Aave's WETH markets was not abstract. Stablecoin and ETH yields on Aave have functioned as a practical savings tool in markets where formal banking access is uneven. Funds deposited in affected pools could not be moved during the crisis. The 16% drop in the AAVE token price, to around $92, compounded losses for holders tracking valuations in local currencies. In Nigeria specifically, AAVE fell 13.99% in Naira terms during this period, a concrete measure of the damage in one of DeFi's most active emerging markets.
The incident also sharpens a structural concern running through all of these communities. The 130,000 delegated AAVE votes going to an Asia-linked institutional treasury is a direct instance of the governance asymmetry that African and South Asian user communities face: they lack treasury-backed representation in DeFi governance despite being among the most active participants in the protocols they use. Decisions affecting their funds are made by large, well-capitalized entities with institutional stakes in the outcome, while the users most dependent on these protocols have no comparable seat at the table.
What comes next
Mantle is not acting alone. A coalition called DeFi United, coordinated by Aave founder Stani Kulechov, has assembled pledges from EtherFi (5,000 ETH), Lido (up to 2,500 stETH), Golem Foundation and Factory combined (1,000 ETH), Frax, Ethena, Ink Foundation, LayerZero, and others. LayerZero's participation is notable given that its bridge infrastructure was the attack vector in the original exploit. Kulechov has committed 5,000 ETH from his own holdings. "Our priority is our users, and every decision we are making is aimed at an orderly return to normal market conditions," he said. Whether the combined effort covers Aave's full exposure depends on which bad debt scenario plays out and how quickly governance on both sides can move.
Aave's treasury holds roughly $181 million in assets and its safety module has about $54 million in staked WETH, putting the combined figure at approximately $235 million. That is broadly comparable to the worst-case bad debt projection of $230.1 million, but the headline comparison overstates the coverage available. The $181 million in treasury assets includes roughly $54 million in AAVE tokens and $52 million in stablecoins, neither of which can directly plug an ETH-denominated shortfall at par. Asset composition, not total size, is the binding constraint, and external capital remains the central question in the weeks ahead.
For users waiting on liquidity normalization, the timeline depends on two separate governance processes moving in sequence. MIP-34 must first pass a Mantle community forum poll and a Snapshot vote before Aave DAO can begin its own independent approval process. Both DAOs must act, and neither can conclude until the other completes its prior step. A swift resolution is unlikely.