The mempool is the one place where Ethereum’s promise of transparency becomes a weapon against its users. Every pending transaction sits in plain text, waiting for a bot to front-run, sandwich, or back-run it. On August 19, a group of Ethereum researchers convened under the banner “Encrypt the Mempool” to discuss a solution that has been simmering in EIP drafts for months: hiding trade contents from bots before they can attack. The trigger for this renewed urgency? A transaction from Vitalik Buterin’s address was exploited by Jaredfromsubway.eth, a notorious MEV bot. The irony is brutal—the man who designed the network cannot safely transact on it.
But this is not a story about a single hack. It is about a structural flaw in Ethereum’s transaction ordering layer, and the proposed fixes—LUCID, EIP-8105, FOCIL—that aim to patch it. After spending years analyzing cross-border payment flows and institutional custody patterns, I have learned one thing: macro breaks micro. Always. The mempool problem is a microcosm of a larger failure in permissionless systems: the tension between public visibility and private value. The question is whether the proposed encryption schemes can resolve this without introducing new forms of centralization.
Context: The Mempool as a Public Good Gone Wrong
Since the emergence of Flashbots in 2020, the Ethereum community has been fighting a rear-guard action against MEV (maximal extractable value). Private relays, like those operated by Flashbots, allow users to submit transactions directly to block builders, bypassing the public mempool. But these relays are trusted intermediaries. They can see your transaction, they can censor it, and they create a two-tier system where informed users pay for privacy while everyone else remains exposed.
Jaredfromsubway.eth’s attack on Vitalik’s wallet was a reminder that even the most sophisticated users are vulnerable. According to on-chain data, MEV losses have declined from peaks in 2022, but the absolute numbers remain high—over $400 million extracted in 2024 alone. The current approach—private relays + auction mechanisms—has reduced the problem but not solved it. The Ethereum Foundation’s research group is now pushing for a protocol-level solution that would make the mempool encrypted by default.
Core: The Technical Architecture—LUCID, EIP-8105, and FOCIL
The centerpiece of the proposal is EIP-8184, known as LUCID (Liquidity Unlocking via Confidential In-Order Delivery). LUCID is a commit-reveal scheme: a block builder submits a sealed block without knowing the transaction contents. After the commit, the sender or a designated key publisher releases the decryption key. The block is then unlocked and executed. In theory, this prevents front-running because the builder cannot see the orders until they are committed.
But here is the catch. The cryptographic primitives required to make LUCID work at Ethereum scale—small public keys, non-interactive decryption, no trusted setup, ciphertext sizes that fit within block gas limits, strong chosen-ciphertext security, and a path to post-quantum security—do not exist yet. The EIP-8184 author explicitly states: “No known cryptographic construction satisfies all requirements simultaneously.” This is not a minor detail. It is a fundamental admission that the technology is not ready.
LUCID sidesteps this by outsourcing decryption to a key publisher. The sender encrypts the transaction, the builder includes it in a sealed block, and then the key publisher releases the secret. This shifts the trust assumption from the block builder to the key publisher. But the key publisher is a new trust anchor—centralized, potentially colluding with the builder, or simply failing to reveal the key on time. The protocol imposes a penalty: a reserve fee equal to 1/8 of the block gas limit, which is forfeited if the key is not revealed. But this fee can also punish honest senders if their key publisher is offline or malicious. The reserve fee is a blunt instrument, not a robust economic incentive.
EIP-8105 proposes a different approach: a directed trust graph. Each provider (e.g., wallet, relay, builder) registers which other providers they trust. Transactions are routed through chains of trust, and the protocol does not enforce any penalty for misbehavior—that is left to off-chain reputation systems. This is flexible but fragile. It relies on social coordination and does not solve the problem of proving bad behavior on-chain.
FOCIL (EIP-7805) is the inclusion list pipeline that ties these proposals together. It allows multiple validators to specify which transactions a block builder must include. This is intended to prevent censorship and ensure that encrypted transactions cannot be simply ignored. FOCIL is already listed as a consensus-layer priority for the Hegotá upgrade, scheduled for 2027. That means the earliest deployment of any integrated encrypted mempool is at least two years away, and likely longer given the unresolved cryptographic challenges.
From my experience modeling DeFi liquidation cascades during the 2020 liquidity mirage, I have seen how fragile trust assumptions become under stress. The LUCID key publisher is a single point of failure. If the publisher is compromised, the entire block’s transactions are exposed. If the publisher goes offline, the reserve fee is burned—a loss for the sender. The economic model is not self-consistent. The reserve fee is meant to deter spam, but it also creates a disincentive for legitimate users who might experience a failed reveal.
Contrarian: The Real Goal Is Not Eliminating MEV—It’s Raising the Cost of Attack
The marketing around encrypted mempools paints a picture of a fair, front-running-free Ethereum. But the technical reality is different. The 1/8 gas limit on the encrypted block top segment means that only a fraction of transactions can be protected. The rest remain in the public mempool. The reserve fee mechanism makes it expensive to attack, but it also makes it expensive to use. The net effect is not the elimination of MEV, but a shift in the cost-benefit calculation for searchers. The most profitable attacks—those targeting large, high-value swaps—will still be possible if the attacker can pay the fee. The smallest users will be priced out of the encrypted zone, effectively creating a two-tier system based on ability to pay the reserve fee.
Furthermore, the trust graph of EIP-8105 does not solve the problem of collusion. If a group of providers agree to front-run each other’s transactions, the graph can propagate that behavior. Off-chain reputation systems are notoriously slow to react and can be gamed. The same problem exists in the private relay market today: Flashbots relays are trusted, but they could theoretically collude with builders. The encrypted mempool proposals do not eliminate this trust; they just move it to a different set of actors.
Another blind spot is the quantum threat. The roadmap explicitly requires a “credible path to post-quantum security” because many of the encryption schemes under consideration (e.g., pairing-based cryptography) are vulnerable to Shor’s algorithm. With Q-Day potentially arriving as early as 2029, any encryption scheme deployed in 2027 must be upgradeable. The current proposals do not specify how such an upgrade would occur without a hard fork. This is a structural risk that could lead to a rushed migration or a permanent security gap.
Takeaway: Positioning for the Long Cycle
The encrypted mempool is not a catalyst for short-term price action. It is a long-term infrastructure bet that will not materialize until at least 2027. For now, the market should treat this as a narrative shift—the Ethereum community is acknowledging that private relays are not enough, but the solution is not yet viable. The immediate impact will be a continued reliance on existing MEV mitigation tools: Flashbots, private order flow, and cross-chain bridges that bypass the Ethereum mempool entirely.
From an institutional flow perspective, the key data to watch is the number of validators adopting FOCIL and the development of post-quantum cryptography in the Ethereum research ecosystem. If a credible construction emerges before 2027, the timeline could accelerate. Until then, the encrypted mempool remains a structural promise, not a structural reality.
Macro breaks micro. Always. The mempool is the micro. The macro is the global shift toward permissionless, trust-minimized infrastructure. This proposal is a step in the right direction, but it is not a leap. The question is not whether Ethereum can hide your trades from bots—it is whether the cost of that hiding is worth the security it provides.