How a Repeat Attacker Drained $5.87M From 1inch's Most-Trusted Market Maker Twice

A serial exploiter returned to 1inch's TrustedVolumes resolver infrastructure in May 2026, stealing $5.87M via a public-function abuse that required zero new user approvals — exposing systemic trust model failures in DEX aggregator architecture.

Executive Summary

• TrustedVolumes, 1inch's largest liquidity provider, suffered a $5.87 million exploit on May 7, 2026, with stolen assets comprising 1,291 WETH, 16.9 WBTC, and $1.47M in stablecoins — all rapidly consolidated into ~2,513 ETH.

• The attacker is the same threat actor responsible for the March 2025 1inch Fusion V1 exploit that netted $4.5M from the same counterparty, indicating persistent, targeted reconnaissance of the TrustedVolumes resolver ecosystem.

• The exploit reveals a structural weakness in 1inch's trust delegation architecture: custom RFQ proxy contracts with public privileged functions allowed the attacker to self-authorize as an order signer using nothing more than pre-existing user token approvals.

• End-user funds held in wallets were not directly stolen, but all users who ever granted token approvals to TrustedVolumes resolver routes remain at risk unless they revoke, creating a long-tail liability that security firms estimate affects tens of thousands of wallets.

• The incident arrives in a DeFi ecosystem already shaken by the $292M KelpDAO bridge hack in April 2026, heightening regulatory scrutiny and accelerating pressure on aggregators to implement mandatory resolver auditing and on-chain approval rate-limiting.

Background & Market Context

1inch Network launched in 2019 as a DEX aggregator designed to solve fragmented liquidity across Ethereum's burgeoning decentralized exchange landscape. By routing trades across dozens of liquidity sources simultaneously — Uniswap, SushiSwap, Curve, Balancer, and more than 200 others — 1inch optimizes swap execution to minimize slippage and price impact. Over its lifetime, the protocol has facilitated more than $700 billion in swap volume across 12 blockchain networks, establishing itself as infrastructure-grade middleware that institutional traders and DeFi protocols alike have come to depend on.

The Fusion architecture, introduced in late 2022, represented a significant evolution in how 1inch handled order execution. Rather than routing directly through on-chain AMMs for every trade, Fusion introduced a network of off-chain "resolvers" — specialized market makers who compete to fill user orders at optimal prices. TrustedVolumes emerged as the dominant resolver within this ecosystem, handling a disproportionate share of Fusion order flow and acting as a critical liquidity backstop. This privileged position, combined with the inherently permissioned nature of resolver architecture, made TrustedVolumes both a competitive differentiator for 1inch and a concentrated risk vector that would prove catastrophic when exploited.

The macroeconomic and competitive backdrop matters considerably. By early 2026, 1inch's market position had eroded from its 2021-era dominance. According to data from The Block and Dune Analytics, CowSwap had climbed to approximately 22–26% of Ethereum DEX aggregator market share — a sharp rise powered by its batch-auction MEV-resistant model — while Kyber Network was capturing roughly 30%. 1inch, despite processing more than $15 billion in monthly swap volume as of late 2025, faced mounting user trust questions following repeated security incidents. The May 2026 TrustedVolumes exploit lands against this backdrop of competitive pressure, waning token performance (1INCH trading near $0.15–0.16 with a roughly $220M market cap), and a broader DeFi security crisis that saw more than $300M drained from KelpDAO just two weeks prior.

DeFi's systemic risk profile has evolved alarmingly in 2026. The KelpDAO bridge exploit in April, which stranded wrapped ETH across 20 chains and triggered emergency freezes at SparkLend and Aave V3, demonstrated how interconnected protocol dependencies can amplify single points of failure. The TrustedVolumes exploit, while smaller in absolute dollar terms, carries outsized systemic significance: it demonstrates that even a protocol's privileged market-making infrastructure — the entities entrusted with executing trades on behalf of end users — can be weaponized through unvalidated public functions. The question of how DEX aggregators vet, audit, and monitor their resolver networks has become an existential one.

Key Developments

March 5–6, 2025 — Initial 1inch Fusion V1 Exploit The threat actor now known for the May 2026 attack made their first strike against the 1inch ecosystem in early March 2025. The attacker exploited a buffer overflow vulnerability in Fusion v1's deprecated _settleOrder function, manipulating the interactionLength variable to corrupt calldata and impersonate a legitimate resolver. By crafting a specially formed transaction containing a few wei of value, the attacker executed unauthorized swaps worth $5 million. TrustedVolumes bore the brunt of losses, approximately $4.5 million. Halborn Security published a detailed post-mortem confirming the vulnerability class as a calldata corruption issue in obsolete contract logic.

March 2025 — Negotiation and Partial Fund Return In a relatively rare outcome for DeFi exploits, TrustedVolumes entered direct on-chain negotiations with the attacker. According to reports from BeInCrypto and AInvest, the hacker ultimately returned approximately $5 million in stolen assets, apparently in exchange for a bug bounty arrangement. 1inch published a blog post acknowledging the vulnerability in its resolver contracts and pledged a security review of all resolver infrastructure. The attacker's decision to return funds was interpreted as either a white-hat intervention or a calculated move to avoid law enforcement scrutiny.

March–April 2025 — Audit Commitments and Protocol Reassurances Following the Fusion V1 incident, 1inch issued public statements indicating that the vulnerability was isolated to deprecated contract code and that current Fusion v2 infrastructure was unaffected. The team committed to a comprehensive audit of resolver contract interfaces. However, subsequent events suggest that the audit either did not cover TrustedVolumes' custom RFQ proxy infrastructure or that findings were not acted upon in time.

April 19, 2026 — KelpDAO $292M Bridge Exploit While not directly related to 1inch, the KelpDAO bridge hack — described by CoinDesk as "2026's biggest crypto exploit" — dramatically reset DeFi risk tolerance. SparkLend froze rsETH exposure, Aave V3 paused rsETH markets, and Compound, Euler, and Fluid all moved to contain contagion. The event heightened on-chain security monitoring across the ecosystem and created the security environment in which the TrustedVolumes attack would be detected rapidly.

May 7, 2026 (Early Morning UTC) — TrustedVolumes RFQ Proxy Exploit At approximately 02:00–03:00 UTC on May 7, 2026, the attacker — using address 0xC3EBDdEa4f69df717a8f5c89e7cF20C1c0389100 — began executing the exploit against TrustedVolumes' custom RFQ proxy contract at 0xeEeEEe53033F7227d488ae83a27Bc9A9D5051756, which interfaced with the TrustedVolumes resolver at 0x9bA0CF1588E1DFA905eC948F7FE5104dD40EDa31. The attack exploited an unguarded public function to self-register as an "Allowed Order Signer," then leveraged historical token approvals granted by users to 1inch's Fusion routing contracts. The primary exploit transaction hash is 0xc5c61b3ac39d854773b9dc34bd0cdbc8b5bbf75f18551802a0b5881fcb990513.

May 7, 2026 — Detection and Fund Consolidation Blockaid and PeckShield independently detected the exploit in real-time through anomalous approval-usage pattern monitoring. By the time alerts were published, the attacker had already consolidated stolen assets — 1,291.16 WETH, 16.939 WBTC, 1,268,771 USDC, and 206,282 USDT — into approximately 2,513 ETH. Security researchers immediately identified the attacker's infrastructure as matching the March 2025 exploit operator. GoPlusSecurity issued warnings for users to revoke approvals via tools like Revoke.cash. Total confirmed losses were $5.87 million.

Technical Analysis

The TrustedVolumes exploit belongs to a class of vulnerabilities best described as privileged-function access control failures in resolver proxy architecture. Understanding why this attack worked requires understanding how 1inch's Fusion system distributes trust.

In Fusion's architecture, resolvers are whitelisted market makers who receive signed user orders and fill them at agreed prices. To enable efficient off-chain negotiation and on-chain settlement, resolvers typically deploy custom proxy contracts that sit between the user's signed intent and the 1inch Settlement contract. These proxies are meant to be tightly controlled, since they inherit the ability to execute trades against pre-existing user token approvals. The critical security assumption is that only authorized signers within the resolver's organization can submit orders through these proxies.

TrustedVolumes' custom RFQ proxy contract contained a public-facing function for managing the "Allowed Order Signers" whitelist. In a secure implementation, this function would be protected by an onlyOwner modifier or multi-sig governance. Instead, it was callable by any external address. The attacker exploited this by calling the function with their own address, self-granting order signer privileges. Once registered as a legitimate signer within the proxy's access control structure, the attacker could craft and submit arbitrary orders against the resolver contract, directing fund transfers to addresses they controlled.

The attack's most insidious dimension is its zero-new-approval requirement. Traditional DeFi exploits often leave an on-chain trace in the form of malicious approval transactions that security monitoring can flag. In this case, users had already granted token approvals to 1inch's Fusion routing infrastructure — approvals given legitimately months or years earlier to enable swap execution. The attacker's self-authorization as an order signer was sufficient to weaponize those dormant approvals, transferring tokens from user-associated positions and from the resolver's own liquidity pool simultaneously. Security firms noted that the exploit may have remained active for an extended period, suggesting that additional losses beyond the confirmed $5.87M may exist.

The connection to the March 2025 Fusion V1 attack is technically and strategically significant. The original attack exploited a calldata corruption (buffer overflow) in the _settleOrder function of an obsolete Fusion v1 contract. The 2026 attack exploits an access control failure in a custom RFQ proxy. These are different vulnerability classes in different contracts — but they share a common theme: both arose from 1inch's resolver ecosystem operating custom contract logic that was not held to the same auditing standard as 1inch's core settlement contracts. The attacker appears to have conducted sustained reconnaissance of TrustedVolumes' contract infrastructure over the 14 months between attacks, identifying the RFQ proxy as the next weakest link after the Fusion V1 vulnerability was patched.

The asset conversion pattern post-exploit is consistent with sophisticated operational security. By consolidating WETH, WBTC, and stablecoins into native ETH, the attacker reduced the number of on-chain entities holding identifiable assets, simplified laundering through mixers or cross-chain bridges, and eliminated the risk of USDC/USDT contract-level freezes (which Circle and Tether can execute on specific addresses). Blockchain analytics firms will track the ETH as it moves through Tornado Cash, Railgun, or cross-chain bridge contracts in the days following.

sequenceDiagram
    participant Attacker as Attacker (0xC3EB...)
    participant ProxyContract as TrustedVolumes RFQ Proxy (0xeEeE...)
    participant Resolver as Resolver Contract (0x9bA0...)
    participant UserWallet as User Wallets
    participant Settlement as 1inch Settlement Contract
    participant CEX as ETH Consolidation

    Attacker->>ProxyContract: Call public addOrderSigner(attacker_address)
    Note over ProxyContract: No access control — succeeds
    ProxyContract-->>Attacker: Attacker now registered as Allowed Order Signer

    Attacker->>ProxyContract: Submit crafted RFQ order (signed by self)
    ProxyContract->>Settlement: Forward order as authorized resolver
    Settlement->>UserWallet: Pull tokens via pre-existing approvals
    Settlement->>Resolver: Drain resolver liquidity pool

    Note over Attacker: Assets received: 1,291 WETH + 16.9 WBTC + $1.47M stables
    Attacker->>CEX: Swap all assets → 2,513 ETH
    Note over CEX: Blockaid & PeckShield detect in real-time
    Note over Attacker: Same actor as March 2025 Fusion V1 exploit

On-Chain & Market Data

The market data paints a complex picture of a protocol caught between operational scale and structural fragility. 1inch continues to process enormous notional volume — $700 billion lifetime and $15 billion monthly — but this throughput is the property of its routing algorithm, not assets it holds in custody. The $2.8M TVL figure underscores that 1inch itself is a flow-through infrastructure layer; the risk resides not in the protocol's balance sheet but in the counterparty contracts — like TrustedVolumes — that sit at its edges.

The token performance is perhaps the most telling macro signal. 1INCH trading at $0.15–0.16, down more than 90% from its 2021 all-time high, reflects not just general market conditions but a persistent governance discount applied by the market to protocols that have struggled with repeated security incidents. The March 2025 hack prompted only a temporary recovery after funds were returned; the May 2026 exploit, arriving against a backdrop of already-eroded trust and intensifying competition, is unlikely to generate a similar recovery catalyst unless accompanied by a credible, independently verified security overhaul.

Competitive Landscape

1inch (the subject protocol) occupies an increasingly precarious leadership position. Its Pathfinder routing algorithm remains technically sophisticated, covering 200+ liquidity sources and 12 EVM chains. However, back-to-back resolver exploits — both involving TrustedVolumes — have introduced reputational damage that pure routing performance cannot easily offset. Institutional liquidity providers evaluating resolver partnerships will scrutinize 1inch's audit commitments more aggressively, potentially leading to market maker attrition precisely when the protocol needs them most.

CowSwap / CoW Protocol has emerged as the most philosophically divergent competitor. Its batch-auction architecture with uniform clearing prices eliminates MEV sandwich attacks and removes the resolver-trust problem entirely by not relying on privileged off-chain market makers in the same way. CowSwap's rise to 22–26% market share on Ethereum is directly correlated with DeFi users' growing sophistication about execution quality and security. In the aftermath of the TrustedVolumes exploit, CowSwap is well-positioned to capture users fleeing 1inch's Fusion routes, though its batch-auction model introduces latency that makes it less competitive for time-sensitive large trades.

0x Protocol / Matcha operates as the DeFi ecosystem's most broadly integrated swap API. By powering MetaMask Swaps, Coinbase Wallet, Robinhood Web3, and dozens of institutional fintech applications, 0x has built distribution moats that are extraordinarily difficult to displace. Unlike 1inch's vertically integrated resolver model, 0x's API abstraction layer distributes execution risk across many liquidity sources without privileged resolver contracts, reducing — though not eliminating — the attack surface that claimed TrustedVolumes.

ParaSwap / Velora (rebranded in 2026) targets the institutional and large-trade segment where execution quality at scale dominates. Velora's architecture is designed for predictable execution on large block trades rather than retail optimization. It serves a different user profile than 1inch's retail-centric routing but competes directly for institutional flow. Velora's cleaner governance track record gives it a trust premium with funds and protocols evaluating aggregator integrations in the post-KelpDAO security environment.

The competitive summary: 1inch's technical moat in routing breadth and multi-chain coverage remains real, but the trust moat that DEX aggregators ultimately depend on for market maker recruitment and institutional adoption has been significantly eroded. Each of the major competitors has structural properties that either avoid or mitigate the resolver-trust problem that 1inch has twice failed to solve.

Stakeholder Analysis

Liquidity Providers and Market Makers face the most direct financial exposure. TrustedVolumes sustained losses of $5.87M in the May 2026 incident, on top of the $4.5M from March 2025. Even if a negotiated recovery is attempted again, the pattern of repeat exploitation against the same counterparty signals a fundamental architectural failure rather than an isolated incident. Other resolver operators within the 1inch ecosystem — who also deploy custom proxy contracts — face elevated scrutiny and should expect urgent audit demands from institutional partners.

End Users face a more diffuse but real risk. While wallet balances were not directly drained in the May 2026 attack, users who interacted with 1inch Fusion routes and granted token approvals to resolver-connected contracts hold ongoing liability. The attack demonstrated that dormant approvals — granted years ago and forgotten — can be weaponized without any new user action. This "approval debt" problem is endemic to DeFi but rarely this viscerally demonstrated. Security researchers estimate tens of thousands of wallets carry relevant approvals.

Developers and Protocols Integrating 1inch must now evaluate whether the aggregator's resolver infrastructure constitutes an acceptable dependency. Spark Protocol (the DeFi lending layer affiliated with MakerDAO) and KelpDAO have both had recent adverse security events; for protocols managing hundreds of millions in TVL, routing execution through a resolver ecosystem with demonstrated vulnerabilities introduces counterparty risk that risk committees and technical councils will need to formally assess.

Regulators and Policymakers will view the May 2026 attack as additional ammunition for frameworks mandating audits, operational security standards, and liability disclosure for DeFi protocols that handle user funds indirectly through privileged third-party contractors. The EU's MiCA framework, already in implementation, does not specifically address resolver-trust mechanisms, but the pattern of repeated resolver exploits is precisely the kind of systemic risk narrative that accelerates regulatory attention.

Investors and Token Holders face a binary narrative: either this is a buying opportunity before a protocol that has survived two major incidents finally ships the comprehensive security overhaul the market has been waiting for, or 1INCH's downward trajectory from $0.15 continues as competitive pressure compounds reputational damage. The absence of a clear security roadmap with verifiable milestones makes the former narrative very difficult to sustain.

Risk Assessment

1. Continued Resolver Contract Exploitation — The attacker has demonstrated persistent focus on TrustedVolumes' contract ecosystem across 14 months and two distinct exploit vectors. Unless all custom proxy contracts are audited, deprecated, and replaced with hardened implementations, additional attack surfaces likely exist. Severity: Critical. Probability: High, given the attacker's established pattern and the complexity of the resolver contract landscape.

2. Systemic User Approval Liability — The attack's abuse of pre-existing token approvals reveals a structural property of ERC-20 approval mechanics that no single protocol can fully mitigate without user action. Tens of thousands of wallets may carry approvals enabling future exploits via any compromised resolver. A mass-revocation campaign is practically infeasible at scale without protocol-level tooling. Severity: High. Probability: Medium-High, as dormant approvals are difficult to audit and users rarely revoke them voluntarily.

3. Market Maker Exodus from Fusion — If major resolver operators conclude that 1inch's Fusion infrastructure carries unacceptable exploit risk — whether due to reputational, financial, or legal concerns — a deterioration of resolver liquidity would compress fill rates and worsen execution quality for users. A self-reinforcing cycle: worse fills → users migrate to CowSwap or 0x → reduced volume → less revenue → less capital for security improvements. Severity: High. Probability: Medium, contingent on whether 1inch's response is perceived as credible.

4. Regulatory Enforcement Action — Repeated security failures at a systemically important DeFi protocol, combined with the broader 2026 DeFi security crisis (KelpDAO, TrustedVolumes within weeks of each other), create conditions for regulators in the EU, UK, or US to pursue enforcement actions or accelerate rulemaking that imposes liability on aggregator protocols for resolver counterparty risk. Severity: Medium-High. Probability: Low-Medium in a 12-month window, but elevated relative to pre-2026 baseline.

Investment & Strategic Implications

For funds and institutional allocators holding 1INCH or evaluating DeFi aggregator exposure, the May 2026 exploit represents a stress test of management credibility. The March 2025 incident could be explained away as a legacy contract vulnerability in deprecated code — regrettable but bounded. A second attack by the same actor 14 months later, against a different custom contract in the same resolver's infrastructure, is not explicable by bad luck. It suggests that 1inch's security posture does not yet include mandatory auditing of resolver partner contracts as a condition of Fusion participation. Until that changes — verifiably, with published audit reports — institutional allocators should apply a meaningful discount to any 1INCH position thesis that depends on Fusion growth.

For DeFi protocols evaluating aggregator integrations — particularly lending platforms like Spark Protocol that route liquidations or collateral swaps through DEX aggregators — the TrustedVolumes incident argues strongly for multi-aggregator routing with automatic failover. No single aggregator should be a single point of failure for a lending protocol's collateral management. The cost of routing complexity is substantially lower than the cost of being collateralized by stale routes when an aggregator's resolver ecosystem is compromised mid-liquidation. Protocols should also urgently audit their own approval relationships with 1inch resolver contracts and revoke where possible.

For builders in the DEX aggregator space, the TrustedVolumes exploit is an architectural case study in why privileged contract functions — particularly those managing access control — must be treated as critical attack surfaces regardless of how they are used in practice. The "Allowed Order Signer" function was presumably public for operational convenience: easy rotation of signing keys without complex governance overhead. That convenience created a $5.87M liability. The design lesson is clear: any function that can grant execution authority over user-approved assets must be protected by multi-sig governance, time-locks, or guardian circuits. The engineering cost of this is days; the liability cost of skipping it has now been demonstrated twice.

Outlook: 30 / 180 / 365 Days

• 30 days: 1inch will publish a post-mortem and announce a mandatory audit program for all Fusion resolver proxy contracts. TrustedVolumes will likely attempt negotiation with the attacker as it did in March 2025, with a 40–50% probability of partial fund recovery. 1INCH token will underperform the broader DeFi index by 15–25% as market makers reassess Fusion participation. On-chain data will show measurable volume migration toward CowSwap and Matcha.

• 180 days: If 1inch ships a credible resolver security framework — including mandatory proxy contract audits, public resolver transparency reports, and an on-chain guardian circuit for Fusion routes — it has a viable recovery path. Conditional on that: Fusion volume stabilizes and the protocol retains its top-three DEX aggregator position globally. If the security framework is delayed or perceived as superficial, resolver attrition accelerates and CowSwap consolidates its challenge to 1inch for Ethereum market share leadership.

• 365 days: The TrustedVolumes incident will be viewed as either the inflection point at which 1inch meaningfully overhauled its resolver trust model — enabling a more defensible competitive position in a security-conscious 2027 DeFi market — or as the second in a series of exploits that accelerated the protocol's structural decline relative to architecturally safer alternatives. The structural shift in user expectations post-2026's DeFi security crisis (KelpDAO, TrustedVolumes, and other incidents) will likely permanently elevate auditing and transparency standards for aggregator resolver networks, reshaping competitive moats across the entire sector.

References

1. Cryip.co — "TrustedVolumes Exploited for $5.87 Million in Ethereum Hack as Repeat Attacker Returns": https://cryip.co/trustedvolumes-exploited-5-87-million-1inch-resolver-same-attacker/

2. EdaFace News — "1inch Liquidity Provider Trusted Volumes Exploited for $5.87 Million" (May 7, 2026): https://news.edaface.com/2026/05/07/1inch-liquidity-provider-trusted-volumes-exploited-for-5-87-million/

3. Halborn — "Explained: The 1inch Hack (March 2025)": https://www.halborn.com/blog/post/explained-the-1inch-hack-march-2025

4. Cryptopolitan — "1Inch market maker Trusted Volumes got hacked for $4.5M": https://www.cryptopolitan.com/1inch-trusted-volumes-got-hacked-for-4-5m/

5. BeInCrypto — "1inch Hacker Returns $5 Million Stolen Funds After Negotiation": https://beincrypto.com/1inch-hacker-returns-stolen-funds-defi-security/

6. 1inch Blog — "Vulnerability discovered in resolver contract": https://blog.1inch.com/vulnerability-discovered-in-resolver-contract/

7. The Block — "CoW Swap surges to 26% market share, challenging 1inch for DEX aggregator dominance": https://www.theblock.co/post/343583/cow-swap-surges-to-26-market-share-challenging-1inch-for-dex-aggregator-dominance

8. CoinDesk — "The $292 million Kelp exploit: how it happened, and what it means for DeFi" (April 19, 2026): https://www.coindesk.com/business/2026/04/19/the-usd292-million-kelp-exploit-how-it-happened-and-what-it-means-for-defi

9. DefiLlama — "1inch TVL Stats & Charts": https://defillama.com/protocol/1inch

10. Messari — "1inch Network Fundamentals & Overview": https://messari.io/project/1inch-network/protocols/1inch-network

11. Coinspect — "1inch Calldata Corruption (Learn EVM Attacks)": https://www.coinspect.com/learn-evm-attacks/cases/1inch-calldata-corruption/

12. AInvest — "1inch hacker returns 5 million after exploiting 2 year old vulnerability": https://www.ainvest.com/news/1inch-hacker-returns-5-million-exploiting-2-year-vulnerability-causing-stock-surge-2503/