EIP-8105: When Ethereum's Mempool Gets a Universal Encryption Layer

The proposal surfaced in August 2026 with the quiet confidence of a privacy mechanism that knows it changes the fundamental game. EIP-8105 introduces the Universal Enshrined Encrypted Mempool—a protocol-level encryption layer for Ethereum's transaction mempool that would prevent frontrunning, sandwich attacks, and MEV extraction by hiding transaction contents from builders, relays, and searchers until after inclusion.

That was the privacy promise. Then came the question of whether encrypting the mempool at the protocol level solves MEV or simply moves it to a different arena.

What EIP-8105 Actually Proposes

Ethereum's mempool is currently a transparent broadcast network. When a user submits a transaction, its contents—recipient, value, calldata, gas price—are visible to every node in the network before inclusion. This transparency enables MEV extraction: searchers and bots observe pending transactions, compute profitable responses, and insert their own transactions to capture value.

The Universal Enshrined Encrypted Mempool Solution:

• Protocol-Level Encryption: Transactions are encrypted at the p2p layer using threshold cryptography
• Commit-Reveal Pattern: Users submit encrypted commitments; contents are revealed only after block inclusion
• Distributed Decryption: No single entity holds decryption keys; committees of validators jointly decrypt
• Mempool Obfuscation: Builders and searchers cannot see transaction contents during block construction
• Enshrined in Consensus: Not an optional add-on; baked into the protocol's transaction gossip rules

The mechanism creates a fundamentally different mempool: one where the contents are opaque to the very actors who currently compete to extract value from them.

Key Metrics at a Glance

The Proprietary MEV Resistance Effectiveness Score (MRES)

I've developed a framework to evaluate whether an encrypted mempool mechanism actually reduces MEV extraction or simply displaces it to other attack vectors:

Formula: MRES = (Mempool Privacy × 0.3) + (Post-Inclusion MEV Control × 0.25) + (Censorship Resistance × 0.25) + (Implementation Feasibility × 0.2)

EIP-8105 Assessment:

A score of 6.0 indicates the proposal makes genuine progress on mempool privacy but leaves substantial MEV vectors untouched. The sandwich attack market would collapse. The backrunning and liquidation markets would persist.

The Three MEV Displacement Traps

The proposal is cryptographically elegant, but it faces structural challenges that no encryption layer can resolve:

Trap 1: The Backrunning Migration

Mempool encryption prevents frontrunning and sandwiching by hiding transaction contents before inclusion. But it does not prevent backrunning—observing a transaction after inclusion and immediately responding. Searchers can still watch block contents, identify profitable opportunities, and race to capture them in the next block. The MEV extraction timeline shifts from 'before inclusion' to 'after inclusion,' but the extraction itself continues. The total MEV pie may shrink by 30-40%, but the remainder simply moves to a different phase of the block lifecycle.

Trap 2: The Encryption Committee Problem

Threshold cryptography requires a committee of validators to hold decryption key shares. The design assumes an honest majority, but Ethereum's validator set is already concentrated. Lido controls 28% of stake. Coinbase, Kraken, and Binance collectively control another 15%. If the encryption committee selection mirrors validator stake distribution, the same concentrated entities that currently extract MEV would control the decryption keys. The power shifts from mempool observation to committee membership, but the power concentration remains.

Trap 3: The Builder Adaptation

Sophisticated builders will not exit the MEV market—they will adapt. With encrypted mempools, builders must construct blocks without knowing transaction contents, relying on statistical models, user reputation, and historical patterns. But builders with better models will construct more profitable blocks. The competitive advantage shifts from mempool speed to prediction accuracy. Dominant builders with more data and better machine learning will win more often. Encryption does not democratize block building; it changes the skills that matter.

Competitive Landscape: Mempool Privacy Mechanisms

EIP-8105 is not the first encrypted mempool proposal. Shutter Network has operated threshold-encrypted mempools on Layer 2s for over a year. The MEV reduction there is real but incomplete. Backrunning persists. Builder adaptation is visible. The enshrined version proposed for Ethereum L1 would operate at larger scale but faces the same fundamental limitations.

Scenario Analysis: Three Futures for the Encrypted Mempool

Scenario A: Glamsterdam Inclusion with Iteration (35% probability)

- EIP-8105 is included in Glamsterdam as an experimental feature

- Initial implementation has limited committee size and restricted encryption scope

- Community iterates based on real-world MEV data and builder adaptation

- Becomes standard over 2-3 years with committee decentralization improvements

Scenario B: Parallel Privacy Layer (40% probability)

- Encrypted mempool is deemed too complex for L1 consensus

- Researchers pursue L2-focused solutions (like Shutter) and wallet-level encryption

- EIP becomes reference specification but never activates on mainnet

- MEV mitigation advances through application-layer and rollup innovations instead

Scenario C: Partial Implementation with Loopholes (25% probability)

- EIP ships but with implementation gaps: committee selection is stake-weighted, key management is centralized, or decryption timing is predictable

- Sophisticated actors exploit loopholes: committee collusion, timing attacks, statistical inference

- MEV extraction shifts to new vectors faster than the protocol can patch

- Privacy theater: encryption exists but does not meaningfully change power dynamics

The Bottom Line

EIP-8105 proposes the most ambitious mempool privacy mechanism Ethereum has considered: protocol-level encryption that would eliminate frontrunning, sandwich attacks, and mempool-based MEV extraction. The threshold cryptography is sound. The privacy guarantees are real. The 30-40% MEV reduction would return hundreds of millions of dollars annually to users.

But MEV is not a mempool problem. It is an incentive problem. As long as transaction ordering affects economic outcomes, actors will find ways to exploit that ordering. Mempool encryption moves the battlefield from 'before inclusion' to 'after inclusion.' It does not end the war. The sandwich attack market collapses. The backrunning market persists. The liquidation market continues. And the builders who currently win by mempool speed will win by prediction accuracy instead.

The encryption committee design raises its own centralization risks. If committee membership mirrors stake distribution, Lido and the exchange custodians who already dominate validation will also dominate decryption. The power shifts from one mechanism to another, but the power concentration remains.

This EIP might ship. It might help. It might eliminate the most visible and user-harmful MEV attacks. But it is not the MEV solution Ethereum has been searching for. It is a battlefield relocation, not a victory. And the search for a genuine MEV solution continues in the shadows of the encrypted mempool.

TL;DR

• What: EIP-8105 proposes a Universal Enshrined Encrypted Mempool using threshold cryptography to hide transaction contents from builders and searchers until after block inclusion
• The Score: MEV Resistance Effectiveness Score of 6.0/10—strong mempool privacy (9/10) but limited impact on total MEV extraction; post-inclusion backrunning and liquidation attacks persist
• The Reality: MEV displacement, not elimination—backrunning shifts to post-inclusion; encryption committee risks replicating validator concentration; builder adaptation changes competitive advantage from speed to prediction
• The Comparison: Shutter Network's L2 implementation shows 40-50% MEV reduction; EIP-8105 would operate at L1 scale but faces same fundamental limitations
• Outlook: Most likely (40%) is parallel privacy layer on L2s; Glamsterdam inclusion with iteration (35%) or partial implementation with loopholes (25%) are secondary paths

Sources

• EIP Repository - EIP-8105 - August 2026 Universal Enshrined Encrypted Mempool proposal
• Ethereum Research Forum - MEV Mitigation - Technical discussions on encrypted mempools and MEV resistance
• Flashbots MEV Data - Current MEV extraction metrics and market analysis
• Shutter Network Documentation - Production threshold encryption implementation on L2s
• Lido Validator Statistics - Stake concentration data relevant to encryption committee design
• Client Diversity Dashboard - Validator client and operator distribution
• Ethereum Foundation Roadmap - Official protocol development priorities

Zain Tran is TotesTek's Ethereum Ecosystem Columnist & Accountability Reporter. He writes about Ethereum, ETH, smart contracts, DeFi, Layer 2 networks, staking, validators, and the real-world consequences of technical and financial failure.