LayerZero says Kelp’s own setup opened the door for a $290M exploit — and points a finger, with preliminary confidence, at North Korea’s Lazarus Group. What happened LayerZero says the April 18 attack didn’t exploit protocol code but instead targeted infrastructure: two remote procedure call (RPC) nodes that LayerZero’s verifier used to confirm cross-chain messages. RPCs are the servers applications use to read and write data on a blockchain. According to LayerZero, attackers replaced the software binaries on two of those nodes with malicious versions that selectively lied to LayerZero’s verifier while continuing to serve accurate data to all other clients. That selective deception kept the tampering hidden from LayerZero’s monitoring, which queries the same RPCs from different IPs. Complicating the attack, the perpetrators also launched a distributed denial-of-service (DDoS) against other, uncompromised external RPC nodes between 10:20 a.m. and 11:40 a.m. PT, causing the verifier to fail over to the poisoned nodes. Once the failover occurred, the compromised nodes fed a forged but apparently valid cross-chain message to LayerZero’s verifier and Kelp’s bridge released 116,500 rsETH to the attackers. The malicious node software then self-destructed, wiping binaries and local logs. Why LayerZero blames Kelp’s configuration LayerZero says the exploit only succeeded because Kelp ran a 1-of-1 (single-verifier) configuration for the rsETH bridge — meaning LayerZero Labs was the sole party signing and verifying messages for that integration. LayerZero’s public integration checklist and direct communications had recommended a multi-verifier setup, where multiple independent verifiers must reach consensus before a cross-chain message is accepted. In a multi-verifier design, poisoning a single verifier’s data feed would not have been sufficient to forge a valid message. LayerZero says it has confirmed no contagion to other apps on the protocol: all OFT-standard tokens and applications using multi-verifier setups were unaffected. The LayerZero Labs verifier is back online, and the company announced it will no longer sign messages for any application running a 1-of-1 configuration, effectively forcing integrators to migrate off single-verifier designs. Attribution and context LayerZero attributed the attack, with preliminary confidence, to the Lazarus Group — specifically its TraderTraitor subunit. If accurate, this would tie the Kelp incident to another recent Lazarus-linked exploit: the April 1 Drift Protocol breach. Together, those two incidents account for more than $575 million drained from DeFi in roughly 18 days, using two different attack vectors — social engineering of governance signers at Drift and infrastructure/RPC poisoning at Kelp — underscoring how quickly sophisticated actors are evolving their methods. What’s next LayerZero’s framing shifts the narrative from a protocol-wide vulnerability to an integrator configuration failure exploited at the infrastructure layer. That distinction matters for how DeFi projects price and manage LayerZero risk going forward. Kelp has not yet publicly responded to LayerZero’s account or explained why it ran a 1-of-1 verifier despite explicit recommendations to use redundancy. Bottom line The Kelp exploit is a reminder that cross-chain security isn’t only about smart-contract code — it’s also about infrastructure and operational choices. Protocols and integrators will likely accelerate moves to multi-verifier architectures and hardened RPC setups as teams race to close the kinds of gaps attackers are increasingly exploiting. Read more AI-generated news on: undefined/news