DeFi protocols have lost more than $771 million to exploits in the first four and a half months of 2026. That figure is on track to exceed every full-year total since 2023 with eight months still left on the calendar. Two attacks alone — Drift Protocol on April 1 and Kelp DAO on April 18 — account for $577 million of the damage. Everything else is the long tail.

The latest additions to that long tail came this week. On April 29, Syndicate’s Commons bridge was exploited for roughly 18.5 million SYND tokens, which the attacker dumped on the open market for an estimated $330,000 to $400,000 before bridging proceeds to Ethereum. The SYND token cratered 35% within hours. A second incident on Sui — initially reported as an Aftermath Finance perpetuals exploit by CertiK Alert, with the broader Sui DeFi ecosystem suffering a string of compromises in recent days including Volo Protocol’s $3.5 million breach on April 22 and Scallop’s $142,000 flash loan drain on April 26 — pushed the monthly tally even higher.

These are not unrelated events. The pattern across 2026 points to something more structural than a sequence of individual coding errors.

The numbers that actually matter

April 2026 alone produced $606.2 million in DeFi losses across 18 days, according to data compiled by Phemex. That single month was 3.7 times larger than the entirety of Q1 2026 ($165.5 million across 34 incidents from January through March). It made April the worst month for crypto theft since the Bybit breach of February 2025.

Year-to-date losses across 47 incidents now stand at $771.8 million, per the same dataset. For comparison, the first four and a half months of 2025 saw 28 incidents totaling roughly $1.75 billion — but that figure was inflated by the single $1.5 billion Bybit hack. Strip out the Bybit outlier and 2026 is running 68% ahead of 2025 in incident frequency, with attacks averaging one every 2.9 days.

The historical context makes the trajectory clearer. According to Chainalysis data covering prior years, total stolen funds across the entire crypto industry came to roughly $1.7 billion in 2023 and $2.2 billion in 2024 (excluding the Bybit anomaly). DeFi-specific losses have ranged between $1.5 billion and $2.4 billion annually since 2021, with bridge exploits consistently producing the largest single-day numbers in any given year.

Four months into 2026, the industry has already burned through what used to be a typical half-year DeFi loss budget. And the composition of those losses has shifted in ways that should concern anyone holding wrapped or restaked assets.

Two attacks, one weakness

Drift Protocol’s $285 million loss on April 1 was not technically a smart contract exploit. Security firm TRM Labs traced the attack to UNC4736, a North Korean state-sponsored group that spent roughly six months running a social engineering campaign against members of the Drift team. The attackers compromised an admin key, then used a fake token called CarbonVote (CVT) — seeded with $500 of Raydium liquidity and wash-traded into a near-$1 oracle price — to manipulate Drift’s pricing feeds and drain vaults.

The Drift contracts had passed two separate audits, including one from ClawSecure as recently as February 2026. The vulnerability was not in the code. It was in the operational security around the people who held privileged keys.

Kelp DAO’s $292 million loss on April 18 was a smart contract failure, but the technical category is misleading. The bridge that was drained — built on LayerZero’s OFT (Omnichain Fungible Token) standard — held the reserves backing wrapped versions of rsETH on more than 20 different blockchains, including Base, Arbitrum, Linea, Blast, Mantle, and Scroll. The attacker spoofed a cross-chain message that LayerZero’s messaging layer accepted as legitimate. The bridge released 116,500 rsETH — roughly 18% of the token’s total circulating supply — to an attacker-controlled address.

CoinDesk’s analysis identified the underlying flaw as an “upgrade-introduced” vulnerability: a defect that emerged from the interaction between a recent contract upgrade and existing Merkle verification logic. The bug had survived two audits because it only manifested under specific interaction patterns that auditors had not modeled.

What makes Kelp categorically different from Drift is the contagion mechanism. When the bridge reserves were drained, every protocol that had accepted rsETH as collateral on Layer 2s suddenly held tokens with nothing underneath them. Aave faces $123 million to $230 million in potential bad debt, depending on how Kelp ultimately allocates the shortfall, according to a joint report from Aave Labs and LlamaRisk. Aave’s AAVE token dropped 16% in the same session as depositors rushed to withdraw funds. SparkLend and Fluid froze their rsETH markets within hours.

The Bank Policy Institute, in an April 23 analysis, called this “a run on Aave” — language that maps directly onto traditional banking crisis terminology. The trigger was not Aave’s solvency. It was the realization that collateral pledged on a different protocol had become worthless, and that lenders on Aave had no way to know which addresses had been exposed.

The smaller hits tell the same story

The exploits below the headline-grabbing threshold reveal the same underlying pattern.

Rhea Lend was drained for $18.4 million in mid-April; Tether managed to freeze $3.29 million of the proceeds. Chainalysis noted the transaction pattern was consistent with an exit scam, leaving open whether the loss was an external attack or insider extraction.

Volo Protocol on Sui lost $3.5 million on April 22 from three vaults holding WBTC, XAUm, and USDC. The protocol froze remaining funds and recovered approximately $500,000 through coordination with the Sui Foundation. Volo’s TVL had exceeded $2.6 billion at the end of 2025 — meaning the breach hit less than 0.2% of locked assets, but raised broader questions about smart contract auditing standards across the Sui DeFi ecosystem.

CoW Swap lost $1.2 million on April 14 to a domain hijacking attack that gave attackers control over the protocol’s frontend infrastructure. Zerion reported a $100,000 loss to a similar frontend exploit. Purrlend lost $1.5 million on a weekend after a suspicious multisig transaction granted unauthorized bridge access. A flash loan attack on BNB Chain pulled $1.6 million by manipulating reserves on a low-liquidity pool. Silo Finance lost $392,000 to a misconfigured oracle on April 3.

Syndicate’s Commons bridge incident on April 29 — the most recent — looks small in dollar terms but follows the same structural template. The bridge accepted unverified cross-chain messages, the attacker minted SYND tokens out of nothing, and the proceeds were dumped before liquidity providers could react. Syndicate has stated it holds sufficient token reserves to make affected users whole, but the broader pattern is unmistakable: the Commons bridge is the seventh distinct cross-chain bridge exploit of 2026.

Bridge exploits accounted for roughly 47% of total April losses, according to LiveBitcoinNews data. That percentage has been climbing steadily over the past three years. In 2022, when Ronin and Wormhole produced the largest individual losses, bridge exploits made up about 35% of the category. In 2023 and 2024, that share fluctuated between 30% and 40%. The 2026 acceleration is unmistakable.

Why bridges keep breaking

The structural argument, which CoinDesk laid out in detail following the Kelp incident, is that bridges are not really one technology. They are a category of approaches to a problem — moving assets between blockchains that cannot natively communicate — and most of those approaches involve trusted intermediaries, validator committees, or external messaging layers that introduce verification gaps.

A bridge that “locks” assets on Chain A and “mints” wrapped versions on Chain B is functionally a fractional-reserve banking operation: the wrapped tokens are only worth what the locked reserves back. The moment those reserves are compromised, every wrapped instance becomes uncollateralized simultaneously. This is why a single $292 million drain can cascade into hundreds of millions in additional damage at downstream lending protocols.

Native asset bridges like Circle’s CCTP avoid this by burning USDC on the source chain and minting it on the destination, eliminating the need for locked reserves. Intent-based bridges shift the verification burden to competitive solvers who post collateral. Both approaches are growing. But the older lock-and-mint model still dominates the cross-chain economy, and most DeFi protocols built before 2025 still use it.

The auditing problem is harder. Both Drift and Kelp had passed reputable audits before being exploited. The CoinDesk analysis quoted by multiple security firms suggested that “upgrade-introduced” vulnerabilities — defects that only manifest after a contract upgrade interacts with pre-existing logic — have become the dominant exploit category in 2026. Audits typically inspect contracts at a single point in time. They do not always model the full state space of post-upgrade interactions.

The question of coordination

It is worth asking whether the cluster of April attacks is coincidental or coordinated. Security firms are split.

The Drift exploit was definitively attributed to UNC4736, a North Korean group with documented ties to the Lazarus Group cluster. The Kelp DAO exploit was also linked to suspected North Korean actors, according to AInvest reporting on Sui Foundation investigations. Volo Protocol declined to formally attribute its breach but coordinated with the same Sui Foundation infrastructure that was responding to the broader pattern.

If three of April’s top incidents trace back to overlapping threat actors, the implications are uncomfortable. North Korean state-sponsored hacking has historically funded a significant portion of the country’s foreign currency reserves, and the techniques being deployed in 2026 — multi-month social engineering campaigns followed by precision-timed exploits — represent a level of operational sophistication that most DeFi protocols are not equipped to defend against. The 2025 Bybit hack, which moved $1.5 billion through North Korean-controlled wallets, was the largest single theft in crypto history. The 2026 pattern looks like distributed continuation of the same playbook.

Smaller incidents like the Syndicate Commons bridge exploit are unlikely to be state-sponsored. The economics do not work — a $400,000 haul is below the threshold that justifies sophisticated nation-state operational investment. Those attacks are more consistent with opportunistic exploitation by smaller groups or individuals who study published vulnerabilities and look for protocols using similar code patterns. The result is a layered attack environment: nation-state actors targeting the high-value bridges and lending protocols, opportunistic actors picking off the long tail.

What this means for capital at risk

The numbers above describe stolen funds. They do not describe the second-order consequences for DeFi as an asset class, which are starting to look more serious than at any point since 2022.

Aave’s potential bad debt of up to $230 million represents direct contagion from a hack on a different protocol. That is the kind of exposure that prompts professional risk managers to reconsider whether DeFi lending markets can be modeled at all. If a $292 million drain at Kelp can produce $230 million in bad debt at Aave, then the effective leverage in cross-protocol DeFi is higher than the headline TVL numbers suggest.

The Bank Policy Institute’s analysis went further, framing the Kelp-Aave cascade as a structural illustration of why DeFi lending platforms cannot be regulated using the frameworks designed for traditional banking. DeFi protocols rely on third-party data feeds and bridge attestations whose reliability cannot be independently verified by the lender. When that information turns out to be wrong — whether through fraud, exploit, or oracle manipulation — the platform absorbs losses that would, in a traditional banking context, be borne by the deposit insurance system. There is no FDIC equivalent in DeFi. The losses just sit on the protocol until token holders or treasuries absorb them.

This is the reason institutional adoption of DeFi has stalled despite multiple regulatory openings in 2025 and 2026. Funds that were positioned to enter DeFi yield strategies after the Genius Act passed have largely held back, citing exactly this category of risk in private conversations reported by Bloomberg and CoinDesk.

What happens next

Three near-term outcomes are worth tracking.

The first is technical. Native bridge models like CCTP and intent-based architectures are gaining adoption but slowly. The cost of migrating a major DeFi protocol to a new bridge standard is substantial, and most protocols will continue to rely on existing LayerZero, Wormhole, and Multichain-style infrastructure for the next 12 to 18 months. Expect more upgrade-introduced vulnerabilities during that transition period. Each protocol that ships a major contract upgrade in 2026 is expanding the attack surface for the next exploit.

The second is regulatory. The Bank Policy Institute’s framing of the Kelp-Aave cascade as a “DeFi run” is being picked up by Treasury Department officials in early discussions about stablecoin reserve rules and lending protocol oversight. The current administration’s enforcement posture remains crypto-friendly, but the bipartisan shape of post-incident analysis suggests that the 2027 Congress, regardless of which party controls it, will face pressure to introduce structural rules around bridge collateralization and DeFi lending disclosures. Protocols operating in the gray zone today should expect that gray zone to narrow.

The third is market-structural. Holders of wrapped or restaked assets have begun rotating into native deposits on centralized exchanges and into protocols using newer bridge architectures. DefiLlama data showed roughly $553 million in net flows out of Ethereum DeFi and into Solana following the Kelp incident, primarily into protocols perceived as having simpler bridge dependencies. This is not a structural exodus. But it is a measurable shift in how sophisticated DeFi capital is hedging against the bridge risk that has now produced two of the three largest hacks in crypto history.

The 2026 numbers, projected forward, point toward a full-year DeFi loss total in the $2 billion to $2.5 billion range — assuming the current pace continues and no single incident matches the 2025 Bybit outlier. That would make 2026 the worst year for DeFi specifically since 2022, when the Ronin and Wormhole exploits anchored the previous record.

The bear case is darker. The bear case is that 2026 is not the worst year. The bear case is that 2026 is the new baseline. Bridge architecture, restaking complexity, and the proliferation of cross-chain wrapped assets have created an attack surface that grows faster than security tooling can keep up. Each successful exploit teaches the next attacker something new. The Drift social engineering playbook was novel in April. By July, it will be a known technique. The Kelp message-spoofing approach will inform every bridge audit through 2027. The defenders are learning. So are the attackers.

The money keeps moving across chains. So does the risk.

This is news analysis based on data from CoinDesk, Bloomberg, Phemex, Chainalysis, TRM Labs, the Bank Policy Institute, AInvest, Aave Labs and LlamaRisk’s joint incident report, CertiK Alert, LiveBitcoinNews, and on-chain transaction analysis. Loss figures reflect publicly reported numbers as of late April 2026 and may be revised as investigations conclude. This is not financial or security advice. Users with exposure to wrapped or restaked assets should review counterparty risk independently.