One entity builds half of all Ethereum blocks. Vitalik Buterin published a blog post on March 2, 2026 laying out a multi-stage plan to break that concentration before it calcifies into a permanent chokepoint.
The proposal arrives weeks before Ethereum's Glamsterdam upgrade, which will enshrine Proposer-Builder Separation (ePBS) directly into the consensus layer. Buterin argues that ePBS alone is not enough. Without additional guardrails, he says, a handful of sophisticated builders will continue to dominate because they can extract the most value from transaction ordering.
Titan Builds Half the Chain, and That Is the Problem
As of late February 2026, Titan Builder produces roughly 50% of all Ethereum mainnet blocks. BuilderNet, the Flashbots-led decentralized builder network that launched to counter this trend, handles about 28%. The remaining 22% splits across smaller operators like Quasar, Eureka, and the once-dominant beaverbuild.
This level of concentration means a single entity decides which transactions appear in every other block. In theory, that entity could censor specific addresses, reorder transactions to maximize its own profit, or quietly exclude competitors' bundles. The risk is not hypothetical. In October 2024, Titan and beaverbuild together controlled 88.7% of block production, and exclusive order flow deals between builders and applications like Banana Gun have channeled user transactions through private pipelines that bypass the public mempool entirely.
Buterin frames this as Ethereum's "Scourge" problem, one of the six research tracks on the official roadmap. The goal: even if 100% of block building falls into hostile hands, no actor should be able to prevent a valid transaction from landing on-chain.
FOCIL: Sixteen Sentinels Guard Every Block
The first line of defense is FOCIL, or Fork-Choice Enforced Inclusion Lists, formalized as EIP-7805. The mechanism works by randomly selecting 16 attesters each slot. Each attester publishes a list of transactions from the public mempool that must appear in the next block. If the builder ignores those lists, the network rejects the block.
FOCIL is already scheduled as the consensus-layer headline for the Hegota upgrade, targeting the second half of 2026. In its base form, it guarantees that censored transactions can always find a path into a block within one or two slots (24 seconds or less). A hostile builder can still order the MEV-relevant transactions however it likes, but it cannot exclude ordinary user transfers, token approvals, or DeFi interactions.
Ethereum developers have been debating FOCIL since early 2025, and not everyone is comfortable with it. Developer Ameen Soleimani has warned that forcing validators to include all valid transactions, including those from sanctioned addresses, could create legal exposure for node operators in jurisdictions that enforce OFAC compliance. Buterin's post does not address this concern directly, but the design makes censorship technically impossible at the protocol level regardless of any individual participant's preferences.
"Big FOCIL" Shrinks the Builder Role to Near-Irrelevance
The more speculative proposal is what Buterin calls "Big FOCIL," an expansion where inclusion lists grow large enough to cover every transaction in a block. Under this design, the builder's job shrinks to ordering only the small subset of MEV-relevant transactions, things like DEX arbitrage and liquidation bundles, plus computing state roots.
If Big FOCIL works as described, block building becomes a commodity operation. The competitive advantage that lets Titan and beaverbuild dominate today, access to private order flow and sophisticated MEV extraction algorithms, would apply only to a thin slice of each block. The remaining 90% or more of transactions would be dictated by the randomly selected FOCIL committee.
This is a significant architectural shift. Today, builders compete for entire blocks. Under Big FOCIL, they would compete for crumbs. The economic incentive to run a specialized builder operation drops sharply, which is exactly the point.
Encrypted Mempools Kill the Sandwich Before It Starts
The second prong targets toxic MEV directly. When a user submits a swap on Uniswap or any other DEX, that transaction sits in the public mempool before inclusion. Sophisticated actors can see it, front-run it by placing a buy order first, then back-run it by selling after the user's trade executes. This "sandwich attack" costs Ethereum users millions of dollars annually.
Buterin's answer: encrypt transactions so their contents remain hidden until they are finalized in a block. If builders cannot read the transaction, they cannot sandwich it. Several encrypted mempool designs have been proposed over the years, including threshold encryption schemes and trusted execution environments, but none have reached production on Ethereum mainnet.
The timeline for encrypted mempools is less defined than FOCIL. Buterin treats it as a medium-term goal rather than something shipping with Glamsterdam or Hegota.
The Transaction Ingress Layer: Wallets to Blocks
Buterin also flagged a less-discussed attack surface: the path a transaction takes from a user's wallet to the mempool itself. Today, when MetaMask or any other wallet broadcasts a transaction, it passes through RPC nodes that can observe, copy, or redirect it before it reaches the builder. Hostile RPC providers can exploit this visibility for front-running even before the mempool stage.
Proposed defenses include Tor-based routing, custom Ethereum mixnets, and Flashnet-style anonymization layers. These would hide not just the transaction content but also the sender's IP address and the RPC path the transaction travels. Combined with encrypted mempools, this creates a system where no intermediary between the user and the finalized block can read or exploit the transaction.
For users of self-custody crypto cards and DeFi wallets, this matters directly. Every on-chain top-up, every token swap at the point of card funding, is a potential MEV target today. Protocol-level protections would reduce the hidden costs that currently sit on top of disclosed card fees.
The Glamsterdam and Hegota Timeline
The implementation timeline spans two major upgrades:
Glamsterdam (H1 2026): Introduces enshrined Proposer-Builder Separation (ePBS) into Ethereum's consensus layer. This formalizes the builder marketplace that currently runs through third-party relays like MEV-Boost. It does not solve centralization on its own but creates the infrastructure for FOCIL to plug into.
Hegota (H2 2026): Ships FOCIL (EIP-7805) alongside EIP-8141 (native account abstraction). Together, these two EIPs represent the largest single upgrade to Ethereum's transaction lifecycle since the Merge. FOCIL handles inclusion guarantees; EIP-8141 handles smart account capabilities.
Big FOCIL and encrypted mempools sit further out on the roadmap, likely post-2026.
FAQ
What is MEV and why does it cost users money? MEV stands for Maximal Extractable Value. It is the profit that block builders and searchers earn by reordering, inserting, or excluding transactions within a block. Sandwich attacks, front-running, and back-running are the most common forms. These add hidden costs to every DEX swap and DeFi interaction, typically 0.1% to 1% per trade depending on liquidity and size.
Will FOCIL affect Ethereum gas fees? FOCIL adds a small amount of data overhead per block (16 inclusion lists per slot), but the consensus among developers is that this will not meaningfully increase gas costs for end users. The primary effect is on builders, who lose the ability to censor transactions.
When will encrypted mempools go live on Ethereum? No firm date. Buterin's post treats encrypted mempools as a medium-term research goal. Multiple designs exist (threshold encryption, TEE-based approaches), but none have been formally proposed as an EIP for a specific upgrade yet.
Does this affect Layer 2 rollups? Indirectly, yes. L2 sequencers face similar centralization and MEV concerns. If Ethereum's L1 adopts FOCIL and encrypted mempools successfully, L2s may follow the same architectural patterns for their own transaction ordering.
Overview
Vitalik Buterin published a plan to dismantle the block builder oligopoly that currently controls Ethereum transaction ordering. The approach uses three layers: FOCIL forces inclusion of all valid transactions, Big FOCIL reduces the builder role to ordering only MEV-relevant trades, and encrypted mempools eliminate toxic MEV like sandwich attacks by hiding transaction contents until finalization. The first component, FOCIL, ships with the Hegota upgrade in H2 2026. The rest sits on a longer timeline. For Ethereum users funding wallets and crypto cards through on-chain transactions, these changes would reduce the hidden MEV tax that currently inflates every swap.
Recommended Reading
- Vitalik Buterin Says EIP-8141 Will Ship Within a Year, Giving Every Ethereum Wallet the Powers of a Smart Contract
- Ethereum Drops a Seven-Fork Strawmap Through 2029, and Vitalik Wants 2-Second Slots With Quantum Resistance Baked In
- What Happens in the 2 Seconds After You Tap a Crypto Card
