Why does BlackRock value MEV innocence? A trial of "the attacker being attacked."

CN
18 minutes ago
A 25 million dollar on-chain counterattack ultimately got stuck on an old legal question: who made what commitment to whom? Understanding this point is essential to discern what the not guilty verdict protects and what aspects of on-chain behavior it does not absolve.

Written by: Daii

Imagine you are buying something at a small store, someone sees your order first, rushes in to buy the item, and then sells it to you at a slightly higher price. This runner makes money every day by observing others' orders. One day, he sees a familiar large order, as usual, he buys in first, only to find that order did not execute as he expected; the goods remain in his hands, while the real money is taken by the counterpart. This controversy on Ethereum in 2023 has such dramatic twists. But the court’s question is not about who is more likable, but whether that order, that signature, constitutes a legal false promise. (Court ruling, pages 11–16, 34–39)

On November 16, a U.S. federal judge declared brothers Anton and James Peraire-Bueno not guilty. BlackRock previously included their criminal case in its prospectus for its staked Ethereum fund: if MEV is ultimately deemed illegal, the fund's staking business might be disrupted. How did a dispute between on-chain trading bots end up in a Wall Street fund's risk disclosure? The answer lies in a single block proposal on Ethereum: the network currently enters a new time slot every 12 seconds, but not every time slot will produce a block. (Court ruling, pages 1, 4–5, BlackRock staked Ethereum fund prospectus, "MEV" risk paragraph, Official Ethereum block proposal explanation)

1. First, understand the person making money by "queuing"

Ethereum transactions do not mechanically credit in order of button clicks. Transactions must first enter a pending queue, then be placed in a block; who goes first or last can change the prices in a decentralized exchange. MEV, or "Maximal Extractable Value," refers to the additional value participants obtain from including, excluding, and ordering transactions. It has different forms, such as arbitrage and liquidation, and not all MEV should be called attacks; one of the most contentious forms is the "sandwich attack": the bot sees your buy order, buys in ahead to push the price up, lets your transaction execute at a worse price, and then sells for profit afterward. Ordinary traders bear higher transaction costs. (Court ruling, pages 4–5, 11–12, Ethereum Foundation MEV explanation)

Today, professional bots usually hand over transaction combinations of "buy first - sandwich target transaction - then sell" to block builders, which are then relayed through MEV-Boost to validators responsible for proposing blocks. You can think of it as a sealed performance program: the validator first sees the cover of the program, which is the block header, and after signing, the relay reveals the full content. This design is to let builders compete on pricing while preventing validators from pilfering others' ordering schemes before payments or commitments. The bots initially assumed that as long as the whole set of transactions entered the same block in the planned order, the coins they purchased would be sold immediately without bearing long-term holding risks. (Court ruling, pages 7–12, Flashbots MEV-Boost documentation)

To further understand the daily trading scenarios and mechanisms of MEV attacks, you can read my previous article on MEV attacks. What needs to be grasped here is the distinction: the three professional traders referred to as "victim traders" were operating bots to find sandwich opportunities; their losses do not mean that the ordinary users they previously sandwiched did not incur costs. (Court ruling, pages 11–13)

2. How the brothers led the sandwich bots into their trap

Records show that the brothers began studying the behavior of target bots in late 2022, establishing 16 validators and repeatedly testing what kind of transactions could attract them. On April 2, 2023, when the validators they controlled were due to propose a block, they introduced 8 "bait transactions" into the publicly pending queue. The bot saw these transactions as ripe targets for sandwiching and submitted a combination of its buy order, bait transactions, and its sell orders. The prosecution’s initial indictment described the entire process as netting approximately 25 million dollars within about 12 seconds; this amount and narrative belong to the prosecution, not to be construed as the judge’s independent civil accounting for each loss. (Court ruling, pages 12–15, U.S. Department of Justice indictment, pages 8–11)

First, look at what the relay was holding. The block builder provided a complete list of unpublished transactions; the validator initially only saw the block header without transaction content and could only access the list after signing. The block header had two fields that needed to be correctly filled in: the parent block root indicated which previous block the new block followed, and the state root reflected the current on-chain state. Filling in either incorrectly would result in the entire block being rejected by Ethereum. The signature only proves who signed that block header, it does not prove the block header was filled in correctly. If the relay releases the list before confirming the block’s validity, the validator has the opportunity to view others’ transactions and rewrite its block. (Court ruling, pages 7–10, Flashbots incident review)

The vulnerability happened to get stuck at this gate. The open-source program used by the involved Ultrasound relay validated that the signature came from the validator due to propose a block and checked for field existence but did not verify whether the values of the parent block root and state root were correct. Even if it attempted to publish a block that was ultimately rejected by Ethereum nodes, the program still would deliver the complete transaction content to the validators. Flashbots’ incident review classified this as a relay implementation flaw; public information does not indicate why the developers omitted the validity check and cannot simplistically state it was "to speed up” or "intentionally let go." (Court ruling, pages 8–10, 14–16, 38–39, Flashbots incident review)

The brothers filled these two fields, which should have been specific fingerprints, with 0s, and then signed using their real keys. The signature check passed because it indeed matched this "zero-filled block header"; however, this block header described an invalid block that genuine Ethereum nodes would not accept. The relay was yet to release that sealed transaction list. The primary use of filling it with zeros is to obtain private content with a valid signature while ensuring the original block cannot go on-chain. There was no impersonation of someone else's private key; the so-called "fake signature" referred to signing a block header filled with incorrect content. (Court ruling, pages 14–16, 36–39, Flashbots incident review)

The second purpose is to eliminate competition. The original block was rejected, preventing it from competing with the valid block the brothers subsequently submitted for the same time slot. They had already seen the bot's trading arrangements, so they kept the bot's initial buy, substituted in their sell orders, preventing the bot’s final sells from completing as planned. The brothers' validators proposed two blocks for the same time slot and were penalized 1 ETH; the recorded losses for traders in court were about 25 million dollars. Afterwards, Flashbots modified the relay program: if the original block failed to publish, it would no longer release transaction contents to the validators. (Court ruling, pages 15–17, Flashbots incident review and fix, Flashbots fix code)

There was a staggering example during the trial: a bot operator said that during one occasion, their bot was willing to exchange about 2.3 million dollars worth of USDC for SHIB, which at the time was worth only about 8.25 dollars, because the algorithm expected to sell it back at a more favorable price immediately. This was not a SHIB investment that an ordinary person would be willing to hold long-term, but rather an instantaneous transaction predicated on "the next two steps will fully execute." This example illustrates that once a combination trade is broken apart, a seemingly small price differential can turn into a massive exposure. (Court ruling, page 12)

In May 2024, the U.S. prosecution charged the brothers with telecommunications fraud, conspiracy to commit telecommunications fraud, and conspiracy to money laundering, contending that the bait, the invalid block header, and the subsequent transfer of funds constituted a scam. The nearly four-week trial ended in 2025 due to the jury's failure to reach a unanimous decision, resulting in a mistrial. On November 16, 2026, the judge approved the not guilty verdict motion after reviewing trial evidence and ordered the case closed; this decision was based on insufficient evidence, not a jury vote establishing the overall strategy as legitimate. (U.S. Department of Justice 2024 indictment announcement, Court ruling, pages 1, 45–46)

3. The judge is not asking "how significant the loss is," but "where the lie is"

The prosecution proposed two core paths of "lying." First, the bait transactions seemingly promised in the bots’ view that they would be sandwiched as initially planned and included in the final block. The judge did not accept this: a transaction appearing in the public queue merely expresses "I want to perform this transaction," without a commitment to let others sandwich it, and certainly without guaranteeing any combination will definitely go on-chain. The price differentials the bots hoped to earn do not automatically convert into a contractual obligation for the bait party. Especially in an open network where there are no uniform terms of use, one cannot write the expectations of market participants as guarantees spoken by the counterparty. (Court ruling, pages 28–35)

Second, the prosecution stated that signing the block header equates to a commitment to publish the full block as originally intended. The judge also did not accept this: the signature, within the context of the mechanism at the time, identified which validator it was and could subject it to penalties for misconduct; it does not guarantee that the block will be valid or will definitely go on-chain. Moreover, fatal to the prosecution’s case, the relay did not check those two zero-filled fields at all; since the release decision by the receiver does not depend on whether the fields are real, the prosecution has difficulty proving that these erroneous fields represented "material misrepresentations" that caused the counterpart to release transaction contents. The vulnerability was indeed exploited, and 25 million dollars was indeed caught up in the dispute, but telecommunications fraud still must prove its specific statutory elements. (Court ruling, pages 36–40)

The prosecution tried to combine a series of covert actions to argue that the overall situation constituted a scam. The judge acknowledged that the trading strategy might be distasteful but stated that concealing the strategy, exploiting software flaws, and complex post-hoc transfers cannot replace the material misrepresentation required for telecommunications fraud. The money laundering conspiracy charge hinged on proceeds from telecommunications fraud; if the first link is not established, the latter falls as well. The ruling clearly marked boundaries at the beginning: it does not rule on whether actions are ethical, nor does it determine other uncharged crimes or civil disputes. (Court ruling, pages 1, 40–46)

This not guilty verdict's value is primarily that it draws a verifiable line for criminal law: one cannot bypass the proof of "what false representations the defendant made to whom, and whether the counterpart changed actions based on that" simply because on-chain behavior caused massive losses. It also reminds protocol designers that economic penalties, code practices, and legal commitments are three different things. If a crucial step must be trustworthy, software should verify it; if a behavior must be bound by a contract, then terms should be written that can be evidenced as accepted. Not guilty does not mean "any MEV is legal," it merely indicates that this prosecutorial theory and this evidence did not cross the threshold of criminal conviction. (Court ruling, pages 28–46)

4. What BlackRock cares about is whether the profits also go through the same "ordering pipeline"

BlackRock’s iShares Staked Ethereum Trust ETF is not handing fund assets to the brothers. It allows service providers to use the fund's ETH for validation to earn protocol rewards, transaction fees, and possible MEV profits. The prospectus states that staking service providers will include MEV as part of their revenue sources; if MEV is determined to violate applicable laws, the fund's staking plan may be disrupted, and the value of fund shares may be impacted. The prospectus also anticipates that in usual circumstances, approximately 70%–95% of ETH will be staked and discloses that sponsors and service chains collectively obtain 18% of the gross staking revenue, with the remainder staying with the fund. The profits do not appear out of thin air: block ordering has pricing, and the fund is participating in sharing this portion of revenue. (BlackRock staked Ethereum fund prospectus, "MEV," staking arrangements, and revenue distribution sections)

But here, we cannot switch concepts. BlackRock's documents are talking about the MEV-related rewards obtained by validators, not that the fund operates sandwich bots itself, nor that it participated in the transactions that exploited relay vulnerabilities in this case. MEV can come from various trading arrangements; within the same revenue pipeline, there are publicly bid ordering revenues, and there may also be sandwich income where slippage is passed onto ordinary users. The court did not categorize them individually, nor did it grant an overall "license" for the industry. For the fund, this not guilty verdict reduces a threat of broad criminal interpretation but does not eliminate regulatory, contractual, reputational, and specific strategic legal risks. (BlackRock prospectus, "MEV" risk section, Court ruling, pages 1, 45–46)

Another layer of BlackRock’s actions is particularly thought-provoking: its staking agreement requires Coinbase to establish legally binding arrangements with validators and requires validators to commit to controls such as anti-money laundering, sanctions address screening, and regular testing; if validators breach, the fund can request Coinbase to pursue redress. This sharply contrasts with this case: the prosecution failed to prove that the involved validators accepted a specific commitment to "publish exactly as signed," whereas the fund is working diligently to write key obligations into contracts in its business chain. Contracts also cannot guarantee imperfection; the prospectus explicitly states that failures in performance by Coinbase and the validators, recovery costs, and staking losses may still be borne by the fund and holders. (BlackRock prospectus, "Staking Addendum" and "Approved Validator Schedule" provisions)

5. After the not guilty verdict, the real question is who bears the ordering costs

This case leaves readers with a more practical judgment framework. When seeing "staking earnings," first ask whether the revenue comes from prioritized fees voluntarily paid by traders or from cheaper execution prices borne by others; then ask how the profits are distributed among the fund, custodians, and node operators; finally, inquire whether the orderer, if breaching commitments, will rely on on-chain penalties, software validation, or enforceable contracts for accountability. BlackRock's documents at least write the MEV earnings, the 18% fee allocations, and service provider obligations into traceable text, yet still insufficient to inform investors where each MEV truly derives from which trading strategy. (BlackRock prospectus, "MEV," revenue distribution, and staking service terms)

For ordinary on-chain traders, the most direct lesson is not to learn the brothers' techniques. Before trading, check the slippage limits, price impacts, and execution paths; use transaction protections that can reduce the exposure of public pending queues where necessary; but "private sending" still relies on service providers' confidentiality and enforcement commitments and cannot be considered zero-risk insurance. For protocols and institutions, it is more crucial to verify the complete data bound to the signatures, clarify the obligations of block builders and validators, and record who accepted which rules. The market may ultimately choose to strengthen by code, fill gaps with contracts, or have regulations redefine boundaries; this verdict only tells us that we cannot let criminal law solve all missing designs in one go. (Court ruling, pages 7–12, 36–46, Ethereum Foundation MEV explanation, BlackRock prospectus staking service terms)

This is why BlackRock values this case, and the not guilty verdict is still worth a close read by all on-chain participants: a seemingly mere 25 million dollar contention between bots actually pushes "transaction ordering rights" from technical details onto legal and investor ledgers. The court does not give a moral pardon for MEV but presents a proof problem: losses do not equal fraud; gains do not equal no costs. The next real determinant of industry direction will be whether it can clarify the sources of earnings, technical verifications, and accountability.

免责声明:本文章仅代表作者个人观点,不代表本平台的立场和观点。本文章仅供信息分享,不构成对任何人的任何投资建议。用户与作者之间的任何争议,与本平台无关。如网页中刊载的文章或图片涉及侵权,请提供相关的权利证明和身份证明发送邮件到[email protected],本平台相关工作人员将会进行核查。

Share To
APP

X

Telegram

Facebook

Reddit

CopyLink