Cross-chain bridge protocols have lost $340.7 million across 14 exploits in the first five months of 2026, according to data compiled by PeckShield as of June 1. Two incidents alone — the $292 million KelpDAO bridge drain and the $285 million Drift Protocol breach, both attributed to North Korea'...
"We made a mistake by allowing our DVN to act as a 1/1 DVN for high-value transactions. We didn't police what our DVN was securing, which created a risk we simply didn't see. We own that." — LayerZero Labs, KelpDAO Incident Report (May 2026)
Cross-chain bridge protocols have lost $340.7 million across 14 exploits in the first five months of 2026, according to data compiled by PeckShield as of June 1. Two incidents alone — the $292 million KelpDAO bridge drain and the $285 million Drift Protocol breach, both attributed to North Korea's Lazarus Group — account for 76% of all crypto theft this year through April.
The losses are not anomalies. Bridge exploits now represent 68% of all DeFi losses in Q1 2026, up from an already elevated baseline. Since 2022, bridges have accounted for over 69% of total funds stolen in DeFi, with cumulative losses exceeding $2.8 billion. The pattern is structural: bridges concentrate liquidity in smart contracts that lock tokens on one chain and mint equivalents on another, creating honeypots that attract state-level threat actors.
What distinguishes 2026's bridge failures is the attack surface. Neither KelpDAO nor Drift was compromised through a smart contract bug. Both exploits targeted off-chain infrastructure — RPC nodes, oracle feeds, multisig signers, and verification networks — components that sit outside the perimeter of traditional code audits. OpenZeppelin's post-mortem of the KelpDAO incident was titled "$292 Million Lost, Zero Bugs Found." The contracts performed exactly as written. The infrastructure around them did not.
PeckShield's June 1, 2026 security alert documented 14 cross-chain bridge attacks resulting in $340.7 million in combined losses. Individual incidents ranged from $180,000 to $292 million. By mid-May, a separate accounting by ChainUP placed bridge-specific losses at $329 million across eight major incidents.
The broader DeFi security picture is worse. Over $840 million has been drained across more than 50 incidents in five months — a 70% year-over-year increase. Bridge exploits dominate the loss distribution. According to KuCoin's security analysis, combined bridge and sophisticated operations drove over $750 million in total crypto losses by mid-year.
The attackers are not finding new vulnerability classes. The same structural weaknesses — cross-chain message verification failures, compromised key management, and concentrated validator infrastructure — appear repeatedly. What has changed is the scale of capital at risk and the sophistication of the actors targeting it.
On April 18, 2026, attackers drained approximately 116,500 rsETH — roughly 18% of the token's total supply and valued at $292 million — from KelpDAO's bridge built on LayerZero infrastructure. Cybersecurity firms Mandiant and CrowdStrike attributed the attack with high confidence to UNC4899/TraderTraitor, a North Korean state-linked threat group operating under the Lazarus umbrella.
The attack mechanism: This was not a smart contract exploit. The attackers compromised internal RPC nodes that KelpDAO's single LayerZero DVN (Decentralized Verifier Network) relied on to validate cross-chain messages. They simultaneously DDoS'd external nodes to force reliance on the compromised infrastructure. LayerZero's default configuration allowed a 1-of-1 RPC quorum, meaning a single compromised node could authorize fraudulent cross-chain messages.
The verifier attested to a fabricated "burn" event on the source chain. The Ethereum-side contract, receiving what it believed was a legitimate cross-chain message, released 116,500 rsETH to the attacker. The full sequence exploited an operational configuration flaw, not a code vulnerability.
The blame chain: KelpDAO publicly accused LayerZero of setting a "dangerously low" default verification threshold. LayerZero initially deflected, but on May 9, 2026, acknowledged the failure. The vulnerability in the Merkle root verification function had existed since a contract upgrade three weeks before the attack and survived two separate security audits from reputable firms. A bridge contract holding $1.2 billion in TVL, audited twice, contained an exploitable configuration flaw that sat undetected for 21 days.
KelpDAO successfully paused contracts to block a second $95 million theft attempt. The Arbitrum Security Council, coordinating with law enforcement, froze over 30,000 ETH of the attacker's downstream funds.
Seventeen days earlier, on April 1, 2026, attackers drained $285 million from Drift Protocol, the largest decentralized perpetual futures exchange on Solana. TRM Labs and Elliptic attributed the operation to the Lazarus Group.
The attack mechanics: The breach combined social engineering, oracle manipulation, and governance exploitation — all executed in a 12-minute window at the end of a months-long preparation campaign. Attackers social-engineered multisig signers into pre-signing hidden authorizations and exploited a zero-timelock Security Council migration that eliminated the protocol's last line of defense.
The attacker manufactured a fictitious asset — "CarbonVote Token" — with a few thousand dollars in seeded liquidity and wash trading. Drift's oracles treated this fabricated token as legitimate collateral worth hundreds of millions. The attacker then borrowed against this phantom collateral and extracted real assets.
The laundering trail: Assets were consolidated and swapped into USDC and SOL, then partially bridged to Ethereum using Circle's cross-chain protocol. The structured, repeatable laundering flow was designed to obscure origin while maintaining control — a hallmark of DPRK-linked operations. This represents the second-largest exploit in Solana's history, behind only the $326 million Wormhole bridge hack of 2022.
A class action lawsuit was filed on April 15, 2026 by Gibbs Mura, targeting Drift Protocol over the losses.
The KelpDAO exploit demonstrated that bridge risk does not stay contained within the bridge. The attacker used the fraudulently minted rsETH as collateral on Aave, borrowing approximately $236 million worth of WETH against the fake tokens. Because Aave recognized rsETH as a valid collateral asset, the protocol could not distinguish legitimate from exploited tokens in real time.
The bank run: More than $5.4 billion exited Aave in the initial wave as users reduced exposure. Over 48 hours, Aave lost $8.45 billion in deposits, dropping its total value locked from $26.35 billion to $17.9 billion. Across DeFi broadly, aggregate TVL fell from approximately $99.5 billion to $85.21 billion between April 18 and 19 — a $14 billion contraction in two days.
The emergency response: Aave DAO pledged 25,000 ETH to cover the shortfall and prevent bad debt cascade. Aave founder Stani Kulechov personally contributed 5,000 ETH ($8.4 million). Aave formed "DeFi United," a cross-protocol coordination effort to contain the fallout. Analysts at NYDIG placed this incident in the same category as the Ronin ($625 million, 2022) and Nomad ($190 million, 2022) bridge failures in terms of systemic impact.
The 2026 data points to five recurring failure modes:
1. Verification concentration. The KelpDAO incident exposed the risk of single-verifier configurations. LayerZero's default 1-of-1 DVN setup meant one compromised node could authorize fraudulent messages across the entire bridge.
2. Off-chain infrastructure attacks. Neither KelpDAO nor Drift was compromised through a smart contract bug. Both exploits targeted RPC nodes, oracle feeds, signing infrastructure, and operational configurations — components that sit outside the scope of smart contract audits. As OpenZeppelin's analysis concluded: "Smart contract audits don't catch operational risk."
3. Cloud concentration. Nearly 40% of major bridge validators are hosted on the same three cloud service providers. A single infrastructure outage or targeted attack could trap assets or trigger unintended liquidations across multiple protocols simultaneously.
4. Audit limitations. The KelpDAO bridge was audited twice by reputable firms. Both audits missed the configuration vulnerability. The industry's reliance on point-in-time code reviews fails to capture the full attack surface of cross-chain systems, which includes infrastructure, operational procedures, and integration configurations.
5. Composability contagion. When exploited bridge assets enter lending and liquidity protocols, losses cascade. A $292 million bridge exploit became a $14 billion DeFi-wide stress event because of the deep integration between bridge tokens and lending markets.
Manuel Aráoz, co-founder of OpenZeppelin, stated on May 26, 2026 that he now considers "all of DeFi unsafe." His reasoning: "Coding agents are superhuman at finding vulnerabilities, and smart contract security is too asymmetric: defenders need to fix every bug while attackers need just one exploit to steal funds." Aráoz disclosed that he has personally advised friends and family to exit DeFi positions, including in protocols such as Aave, MakerDAO, and Compound.
Less than 2% of DeFi's $83 billion TVL carries any form of insurance coverage. According to DeFiLlama data, 28 insurance protocols operate in the space, but Nexus Mutual accounts for nearly the entire sector's $123.5 million in TVL — 0.14% of DeFi's broader market. Nexus Mutual has paid out more than $18.5 million to cover holders across past exploits, settling 100% of valid claims, often within a week.
However, bridge exploits frequently fall outside typical DeFi insurance coverage boundaries. The KelpDAO incident, which targeted off-chain infrastructure rather than smart contract code, highlighted a gap in existing protection frameworks. The core bridge risk — an operational failure in verification infrastructure — sits in a coverage gray zone.
The economic math is stark: $840 million in DeFi losses in five months against $123.5 million in total insurance TVL. The sector's risk transfer capacity is an order of magnitude below its loss exposure.
The most measurable post-KelpDAO response has been a migration toward Chainlink's Cross-Chain Interoperability Protocol (CCIP). According to Chainlink and CoinDesk reporting, roughly $4 billion in assets shifted to CCIP-connected infrastructure in the weeks following the exploit.
Key migrations include:
CCIP's architecture uses a minimum security threshold of 16 node operators, compared to LayerZero's previous 1-of-1 DVN default. CCIP also implements multiple independent risk management networks and built-in rate-limiting mechanisms designed to prevent massive single-transaction drains.
LayerZero responded by announcing that its DVN will no longer service 1/1 configurations. The company stated it is migrating "all defaults on all pathways" to 5/5 DVN setups where possible and no less than 3/3 on any chain where only three DVNs are available.
The Alephium bridge incident in May — an $815,000 exploit using forged guardian messages that took seven minutes to execute — demonstrated that even smaller bridges with different architectures face similar off-chain infrastructure risks. The exploit involved forged messages that tricked guardians into approving unauthorized transfers, resulting in the minting of 13.76 million unbacked wrapped ALPH on Ethereum.
The 2026 bridge exploit data presents a structural problem, not a cyclical one. The attack surface has shifted from code to infrastructure, but the industry's defense posture — centered on smart contract audits and bug bounties — has not shifted with it.
The economic incentives are clear. Bridges hold concentrated liquidity. State-level actors have the resources and patience to spend months preparing social engineering campaigns and infrastructure compromises. The payoff from a single successful attack ($285–292 million) dwarfs the cost of the operation.
The migration from LayerZero to CCIP addresses one failure mode — verification concentration — but does not resolve the broader structural challenge. Cloud infrastructure concentration, oracle manipulation risk, and the composability contagion problem (where exploited assets cascade through DeFi lending markets) remain unaddressed.
The insurance market offers no backstop. At current capacity, DeFi insurance could absorb less than 15% of year-to-date losses. Until the risk transfer infrastructure scales to match the capital at risk, the economic burden of bridge failures will continue to fall on protocol treasuries, DAOs, and end users.
The data suggests that cross-chain bridges, in their current form, represent the single largest concentration of systemic risk in decentralized finance. The question facing the industry is not whether bridges will continue to be exploited, but whether the structural responses — higher verification thresholds, distributed infrastructure, continuous security monitoring — can narrow the gap between attack capability and defense posture before the next nine-figure incident.