Why Submarine Sends and Covert Channels Are Essential Tools

Public intent broadcast allows bots to insert their own transactions ahead of yours, stealing price improvements. This is a direct, measurable cost extracted from users on every DEX trade.

• Cost: Routinely 5-30+ bps per swap, scaling with size.
• Scope: Affects Uniswap, Curve, Balancer, and all on-chain AMMs.
• Result: Users systematically receive worse prices than the public state suggests.

Transactions are encrypted and sent directly to trusted builders/validators, bypassing the public mempool entirely. This is the foundational privacy primitive for payments.

• Mechanism: Used by Taichi Network, Flashbots Protect, bloXroute Private Txns.
• Benefit: Eliminates front-running and sandwich attacks for simple transfers and swaps.
• Limitation: Relies on trust in the relay not to censor or exploit the transaction itself.

Even private transactions reveal metadata (sender, recipient, approximate value) to relays and block builders, creating new centralization risks and potential for off-chain deal-making.

• Risk: Relays can censor transactions or auction off the right to include them.
• Entity Risk: Creates kingmaker power for entities like Flashbots, bloXroute.
• Outcome: Privacy is not complete; economic security is delegated.

Advanced cryptographic schemes that hide transaction metadata from everyone, including the block builder. The ultimate goal for payment privacy.

• Technology: zk-SNARKs (e.g., Tornado Cash), Oblivious RAM designs, FHE-based L2s.
• Benefit: True sender/receiver/amount privacy, breaking the linkability chain.
• State: Computationally heavy today, but active R&D in projects like Aztec, Fhenix, Penumbra.

Bridging assets reveals intent across chains. Bots can front-run the deposit on the destination chain or perform liability attacks, making bridging risky and expensive.

• Attack Vector: Seen in native bridges and some third-party bridges.
• Consequence: Increases cost and latency for LayerZero, Axelar, Wormhole messages.
• Scale: A multi-billion dollar attack surface as interoperability grows.

Users submit a desired outcome (e.g., "swap 1 ETH for USDC on Arbitrum") rather than a specific transaction path. Solvers compete privately to fulfill it best.

• Protocols: Across, CowSwap, UniswapX use this model.
• Benefit: Hides execution path, creates competition among solvers, and often results in better prices.
• Future: The paradigm shift from transaction broadcasting to declarative finance.

Public mempools expose user intent, allowing searchers to extract ~$1B+ annually in MEV. This is a direct tax on Uniswap, Curve, and Aave users, creating a toxic, adversarial environment for execution.

• Cost: Adds 5-50+ bps to every transaction.
• Risk: Enables sandwich attacks and time-bandit exploits.
• Inefficiency: Forces protocols to build complex workarounds like CowSwap's batch auctions.

Commit to a transaction off-chain, then 'surface' it directly to a block builder, bypassing the public mempool entirely. This is the core mechanism behind intent-based systems like UniswapX.

• MEV Resistance: Transaction is only visible upon inclusion, negating front-running.
• Cost Certainty: User signs with a maximum cost, preventing last-second fee spikes.
• Builder Integration: Direct integration with Flashbots SUAVE or private RPCs like BloxRoute.

Bridging assets via canonical bridges like Wormhole or LayerZero creates a permanent, public link between all your addresses across chains. This enables chain analysis firms to trivially map entire user portfolios and behaviors.

• Data Leak: Source chain, destination chain, amount, and timestamp are all public.
• Protocol Risk: Forces privacy-focused dApps to avoid major liquidity pools.
• User Experience: Kills anonymity sets for protocols like Tornado Cash.

Use a shared liquidity pool or a private settlement layer to break the on-chain link between source and destination transactions. This is critical for privacy-preserving bridges and cross-chain DeFi.

• Link Breaking: Deposit and withdrawal are cryptographically unlinked on public chains.
• Liquidity Efficiency: Uses a shared pool, similar to Across's single-sided liquidity model.
• Composability: Enables private cross-chain swaps, lending, and derivatives.

Relayers, sequencers, and even validators can censor transactions based on origin, destination, or content. This centralizes power and violates the credibly neutral base layer promise of Ethereum and other L1s.

• Single Point of Failure: Relayer-dependent bridges (e.g., some Hyperlane configs) can block TXs.
• Regulatory Risk: OFAC-compliant blockspace threatens DeFi's global access.
• Protocol Risk: dApps can be deplatformed from critical infrastructure.

Distribute the submission or relaying role across a permissionless set of operators, making censorship economically non-viable. This aligns with the security models of EigenLayer and DVT-based validator sets.

• Anti-Censorship: Requires collusion of a large, anonymous operator set.
• Credible Neutrality: Infrastructure cannot discriminate based on transaction content.
• Robustness: Eliminates single points of failure in the transaction supply chain.