Learn how cross-chain bridge exploits work, the most common attack vectors behind major DeFi hacks, and what protocols are doing to harden bridge security.
Cross-chain bridges have transferred hundreds of billions of dollars across blockchains—and lost billions of those dollars to attackers. In a single year (2022), a handful of bridge exploits collectively drained more than a billion dollars from the DeFi ecosystem, making bridge infrastructure one of the most scrutinised and dangerous surfaces in all of crypto.
Understanding how these exploits happen is not just useful for security researchers. It is essential knowledge for anyone deploying capital across chains, evaluating a DeFi protocol, or building Web3 applications. This guide breaks down how bridges work mechanically, why they attract attackers, the most common attack vectors, and what teams are doing to reduce the risk.
A blockchain is, by design, isolated. Bitcoin cannot natively read Ethereum state; Solana cannot verify an Avalanche transaction. Bridges solve this by creating a trust layer between two independent networks.
Most bridges follow one of three architectural patterns.
The user locks an asset (say, ETH) in a smart contract on the source chain. A bridge operator or validator network observes this event, then mints a synthetic “wrapped” version of the asset (wETH) on the destination chain. To redeem, the user burns the wrapped token and the original is unlocked.
The locked asset pool is exactly what makes bridges so valuable to attackers: it is a honeypot of real tokens sitting in a smart contract or controlled by a set of private keys.
Some protocols (such as Hop Protocol or Across) maintain pools of native assets on each supported chain. Users swap rather than lock-and-mint, so the bridge does not hold a single concentrated reserve. This distributes risk but introduces its own complexity around liquidity incentives and price slippage.
More generalised cross-chain messaging protocols (such as Wormhole or LayerZero) do not move tokens directly. Instead, they relay arbitrary messages between chains. Tokens can be part of those messages, but the core product is verified cross-chain communication. These protocols rely on a set of off-chain observers—often called guardians, relayers, or oracles—to attest that an event occurred on the source chain.
Each model has a different trust surface. All of them introduce components that do not exist in a single-chain protocol: multisig key management, off-chain observer networks, and complex state-synchronisation logic.
Three structural factors make bridges disproportionately vulnerable:
Concentrated liquidity. A bridge may hold the canonical reserves for an entire chain’s wrapped assets. Draining it in one transaction can be catastrophically effective.
Cross-domain complexity. Bridge logic spans two or more separate execution environments. A function that is safe on chain A may be exploitable when combined with an event on chain B—an interaction that unit tests for a single chain will never surface.
Multiple trust assumptions. Every component—smart contracts, validator key sets, message relayers, oracle feeds—is an attack vector. A single weak link is enough. Attackers are patient and highly incentivised: the payoff for finding one flaw can exceed the annual revenue of most DeFi protocols.
These dynamics have made bridges a favourite target for sophisticated threat actors, including state-sponsored groups, according to on-chain forensics firms and US government advisories.
Many bridges are ultimately controlled by a small multisig of private keys. The Ronin Network bridge—which backed the Axie Infinity game economy—was drained of approximately $625 million in March 2022 when attackers gained control of five of the nine validator private keys required to authorise withdrawals. Once they held a majority, they issued fraudulent withdrawal transactions and the smart contract complied.
Key compromise can happen through phishing, supply-chain attacks on developer machines, or social engineering. In the Ronin case, investigators including Chainalysis and the US Treasury—which sanctioned an associated wallet address via OFAC—attributed the attack to a North Korea-linked threat group. No smart contract bug was required; the cryptographic governance layer itself was the vulnerability.
Mitigation: Increase the number of required signers, distribute keys geographically and organisationally, use hardware security modules (HSMs), and implement time-locks so that large withdrawals have a mandatory delay before execution.
The Wormhole bridge was exploited for approximately $320 million in February 2022 through a flaw in how the bridge verified cryptographic signatures. The attacker found a way to call a deprecated instruction in the Solana-side contract that allowed them to mint 120,000 wrapped ETH (wETH) without locking any actual ETH on the Ethereum side.
This class of vulnerability—where an attacker forges or bypasses a proof of deposit—is particularly insidious because the bridge logic appears to work correctly. The contract verifies a signature; it just verifies a signature the attacker manufactured.
Mitigation: Exhaustive auditing of all signature-verification code paths, deprecation hygiene (removing old instructions rather than leaving them accessible), and formal verification of cryptographic routines.
The Nomad bridge was exploited for approximately $190 million in August 2022 via a smart contract initialisation error introduced in a routine upgrade. A single storage variable was incorrectly set to zero, which the contract interpreted as a valid “proof of all messages.” This meant any user could copy a previously valid transaction, change the destination address to their own, and resubmit it for a fresh payout.
What made Nomad particularly notable was the attack’s viral nature: once the initial exploit was posted publicly, hundreds of copycat wallets replicated the transaction template, draining the bridge in a chaotic free-for-all rather than a single surgical strike.
Mitigation: Rigorous pre-upgrade testing environments that mirror production state, post-deployment monitoring that alerts on anomalous outflow rates, and circuit breakers that pause the bridge if withdrawal volume exceeds configurable thresholds.
Bridges that rely on external price feeds or off-chain data relayers inherit the attack surface of those systems. If an attacker can manipulate the oracle price used to determine how many tokens to release, or bribe or compromise a subset of relayers to approve a fraudulent message, the bridge contract will execute as designed—on false premises.
This vector is less common for pure asset bridges but is central to cross-chain DeFi protocols that combine bridging with lending, derivatives, or structured products. As platforms like Hyperliquid grow in complexity, the interaction between bridge infrastructure and on-chain financial logic will only increase the severity of potential oracle manipulation.
The bridge security landscape has improved substantially since the exploit-heavy period of 2022, driven by hard lessons, growing bug bounty programs, and institutional demand for safer infrastructure.
Decentralised validator sets. Moving from a small multisig to a large, economically incentivised validator network raises the cost of compromise. Proof-of-stake bonding means a malicious validator risks losing staked capital, not just reputation.
Time-locks and rate limits. Many bridges now enforce mandatory delays on large withdrawals and cap the total value that can leave in a single block or hour. This does not prevent exploits but limits their severity and gives security teams time to pause contracts before full drainage.
Formal verification. Applying mathematical proofs to smart contract logic—rather than relying solely on manual audits—can eliminate entire classes of signature-verification and state-transition bugs. Several bridge teams have contracted formal verification firms for their most critical modules.
Layered auditing. Leading protocols now require multiple independent audits before deployment, maintain continuous audit retainers for upgrades, and run public bug bounties with meaningful payouts. Immunefi is the dominant marketplace for DeFi security bounties.
Insurance and reserves. Some protocols maintain protocol-owned reserves or partner with on-chain insurance markets to backstop losses. This does not prevent hacks but improves recovery prospects for affected users.
Developers who run their own infrastructure—such as those following our guide on running an Avalanche (AVAX) node or selecting RPC node providers—should understand that node-level security is distinct from bridge security, but both contribute to the overall trust assumptions a user accepts when operating cross-chain.
Regulators are paying increasing attention to bridge exploits, particularly when stolen funds flow through sanctioned addresses or are used to finance malicious actors. The US Treasury’s Office of Foreign Assets Control (OFAC) has designated specific mixer services and wallet addresses tied to bridge hack proceeds, creating legal obligations for exchanges and DeFi front-ends to screen for those addresses.
In the European Union, the Markets in Crypto-Assets regulation (MiCA) does not explicitly govern bridge protocols as of its current implementation phase, but its provisions on operational resilience and custody apply to centralised entities that interact with bridge infrastructure. For a full breakdown, see our complete MiCA guide and the MiCA July 1 deadline overview for crypto firms.
In the United States, the SEC has brought enforcement actions touching DeFi in several areas, though bridge-specific enforcement remains nascent. Our SEC Crypto Enforcement 2026 tracker covers the latest actions and their implications for DeFi developers and users.
The broader point: as bridges become systemically important—moving liquidity that underpins liquid staking, cross-chain lending, and interoperable on-chain economies—the political and legal pressure for some form of bridge security standard will intensify.
Last updated: June 2026