Why Optimistic Bridges Are a Subsidy for Attackers

Optimistic bridges are a subsidy for attackers. Their security model externalizes the cost of verification, creating a negative-sum game where honest users pay for the fraud proofs that secure their own funds.

The challenge period is a free option. Attackers can launch unlimited attempts, only paying the gas fee for a failed challenge. This is a classic asymmetric risk, similar to the oracle problem in DeFi lending.

This model inverts economic security. Unlike proof-of-stake bridges like LayerZero or IBC, where capital is slashed, optimistic bridges like Across or Nomad rely on a watchtower economy that often fails to scale with TVL.

Evidence: The Nomad bridge hack exploited a flawed one-block fraud proof window, resulting in a $190M loss. The economic design made a profitable attack inevitable.

The economic model is inverted. Users pay a fixed, small fee for security, while a potential attacker's cost is the entire bonded value. This creates a massive risk asymmetry where the attacker's potential reward (stealing the entire bridge TVL) vastly outweighs their upfront cost.

Honest users are the subsidy. Every transaction fee paid on Across or a similar optimistic bridge funds the liquidity pool that an attacker ultimately targets. Users are paying for a security promise that is only validated after a successful theft is attempted.

Compare this to light clients. Zero-knowledge or validity-proof bridges like zkBridge force attackers to first break cryptographic assumptions, making the cost of attack astronomically high and independent of user fee revenue. The security budget is cryptographic, not economic.

Evidence: The Nomad bridge hack exploited this exact model. Attackers needed only to copy-paste a fraudulent Merkle root, turning a $200 million TVL into a costless bounty funded by user fees. The security delay was a subsidy window.

Optimistic bridges are a subsidy. Protocols like Across and Hop Protocol finalize transfers before verifying them on-chain. This creates a risk window where attackers can steal funds if fraud proofs fail. The protocol's liquidity providers, not the attacker, bear the ultimate loss.

Arbitrageurs are the primary beneficiaries. This design flaw is a free option for MEV bots. They exploit the delay between a user's fast withdrawal and the slow verification on the source chain, capturing value that should secure the system.

The security budget is misallocated. Capital efficiency metrics for bridges like Stargate ignore the real cost of insurance. The liquidity locked for fast withdrawals is a direct subsidy paid by LPs to users and arbitrageurs, not a sustainable security model.

Evidence: The $190M Nomad hack. It was not a sophisticated cryptographic attack. It was a race condition exploit in the fraud proof mechanism, demonstrating how optimistic assumptions fail under adversarial pressure. The bridge's speed became its vulnerability.

The challenge window is a bounty. Optimistic bridges like Across and Hop Protocol require a delay for fraud proofs. This period is a free option for attackers to exploit price discrepancies or chain reorganizations without immediate risk, subsidizing their operations with protocol capital.

Capital efficiency funds the attack. The optimistic security model requires less locked capital than a fully-backed system. This 'efficiency' directly reduces the attacker's cost of capital, lowering the economic barrier to mounting a profitable attack on the bridge's liquidity pools.

The subsidy is quantifiable. The attack ROI is the profit from an exploit minus the capital cost during the challenge period. For a $10M pool with a 7-day window, an attacker needs to lock only the potential penalty, not the full sum, creating a leveraged attack vector absent in ZK-based or native systems.

Evidence: The Nomad bridge hack exploited this model, where a failed fraud proof and low capital requirement led to a $190M loss. This demonstrated that the optimistic delay is not just a feature but a priced-in vulnerability.

The security subsidy is explicit. Optimistic bridges like Across and Nomad rely on a watcher network to flag fraud. This creates a positive expected value (EV) for attackers: the cost of a successful attack is the bond, but the reward is the entire bridge TVL. This is a direct subsidy for sophisticated adversaries.

Watchers are not validators. Unlike proof-of-stake chains where validators slash each other, watchers have no skin in the game. Their incentive is a small bounty for reporting fraud, which fails against zero-day exploits or sophisticated MEV-based attacks that front-run watcher transactions.

The delay is the attack surface. The challenge period (e.g., 20 minutes) is not a security feature; it is the vulnerability window. Protocols like Arbitrum use this model for rollups because the state is verifiable. Bridges move opaque assets, making fraud proofs economically impractical for users to verify.

Evidence: The $190M Nomad hack. The exploit was a public, replayable transaction. The watcher system failed catastrophically because the economic design did not align incentives. Attackers were subsidized by the protocol's own mechanics, turning a bug into a free-for-all.