Diving into Censorship: Ethereum Network

Diving into Censorship: Ethereum Network

Written by
Kiernan Geoghegan
Date published
October 31, 2022

The Ethereum network has a censorship problem. OFAC Compliance, Social Slashing, Centralized Staking Providers, Flashbots, MEV-Boost...These are the most pressing problems to have faced Ethereum, here's everything you need to know about ETH and censorship.ūüßĶūüĎá

image

1/ Buckle up because this is going to be long, and will be broken into a few parts:

A. Background Information

B. Validator Censorship

C. Flashbots Censorship

D. The Future, How to Deal With Censorship Going Forward

2/ A. Background Information

Decentralization and censorship resistance come hand-in-hand. In essence, for something to be decentralized, it must as well be censorship resistant.

3/ Some have argued that this censorship resistance is one of the only real value propositions of blockchains.Without this feature, there is not much to differentiate a blockchain from a cumbersome and laggy database.

4/ On the Ethereum Core Devs call on August 18th, @vdWijden said:"Censorship resistance is the highest goal of Ethereum and the blockchain space in general. If we compromise on that, there's not much else to do"

5/ In the past this threat of censorship has come in somewhat reduced forms.In Ethereum's previous POW model, the only way that censorship could be fully achieved was through something like a 51% attack.

6/ However, a recent series of events has come together in combination to allow censorship to a greater degree on Ethereum.

7/ These events, as follows are:

1. The US Treasury/OFAC sanctioning Tornado Cash and other related addresses on August 8th, 2022.

2. The Merge: Ethereum's switch to a Proof-of-Stake Consensus Mechanism.

3. The use of Flashbot's MEV-Boost middleware by validators.

8/ That first action by the US Treasury has served as a catalyst for the current state that we're facing on Ethereum.For the first time, significant regulatory pressure is being put on companies to censor transactions. Dealing with this could make or break Ethereum.

9/ B. Validator Censorship

As soon as this ruling came out, people immediately looked to the composition of validators that currently made up the network.

10/ Currently, the top staking providers control a significant amount of total staked ETH.

1. Lido at 29.7%

2. Coinbase at 14.3%

3. Kraken at 8.2%

4. Binance at 6.6%

These top 4 parties control a combined 58.8% of all staked ETH.(Retrieved from @hildobby_ staking dashboard)

image

11/ Note: Lido is not equivalent to the other centralized entities here, but is still controlled by the Lido DAO.It is worth adding that Lido token holders' incentives might not necessarily align with those of the greater Ethereum network.

12/ Let's get a little refresher on the requirements for attacks on Ethereum:

33% - Nonfatal; Can delay finality

34% - Nonfatal; Can cause double finality

51% - Debatable fatality; Controls future

66% - Debatable fatality; Controls future and past

13/ As previously mentioned, the top 4 staking providers currently hold a cumulative 58.8% control of the network. That's bad. And it was worse a few weeks ago, the Top 3 were over 66%.But what does this actually mean?

14/ Well it depends on who you ask, as it's pretty debatable.Although at first glance you might see this information and think:"Lido and Coinbase basically control Ethereum, we're f*cked". But it's way more complicated than that, and pretty controversial.

15/ First off, it's unclear what the true legal interpretation of the US Treasury ruling is. Do validators have to stop themselves from building blocks containing transactions from sanctioned parties?Are validators unable to vote for the finality of sanctioned blocks?

16/ A lot of these questions are difficult to answer, and in combination with statements like this from the CEO of @coinbase, @brian_armstrong, many have been left with an uneasy feeling about the future.

image

17/ But wait...There's still hope for Ethereum, but it's going to get a bit controversial.Introducing: Social Slashing and Layer 0

image

18/ I'll be providing a summary here but for a more comprehensive look at the case for social slashing, I highly recommend checking out @ercwl 's medium post on the topic.Additional warning that this is a controversial topic, and that I do support slashing, you may disagree.

19/ Now time for some explanations:

What is slashing?

Slashing is the (sometimes partial) destruction of a validator's stake due to their inability/unwillingness to validate the correct blocks or if they vote to finalize the wrong block.

20/ That's how the process currently functions on POS Ethereum, but a large faction has emerged who believe that slashing should also be used to combat censorship.

If a validator censors tx's, or worse, refuses to attest to a finalized block with sanctioned tx's, slash them.

21/ Now before getting into the obvious implications of a radical change like this let's first address how this could be implemented. Through a User-Activated Soft Fork, or a UASF.

22/ A UASF is essentially a change in the underlying rules that validators adhere to, decided by social consensus.

This may seem like a foreign and outlandish path to take at first glance, but crypto has been here before.

23/ In 2017 a pending Bitcoin Improvement Proposal was unable to garner enough support from miners even though it was supported pretty widely by the community.

The idea eventually came about that change could come about in the form of a User-Activated Soft Fork.

24/ A developer named ShaolinFry explained that soft forks were not actually enforced by miners, but instead by the nodes of the network.In other words, the network's users.

25/ I'll spare most of the details for now, as SegWit and ShaolinFry are both long stories that could get threads of their own.

This was essentially Bitcoin's Independence Day, where the users were able to declare their freedom from the control of the miners.

26/ And now a similar process could be carried out with Ethereum. We can alter the rules for the protocol to punish those that facilitate censorship.

27/ A validator refuses to attest to blocks with sanctioned transactions? They could be slashed.A validator refuses to include sanctioned transactions in the blocks that they've built? They could be slashed. A UASF could put an end to the threat of validator-level censorship.

28/ Now before I get ahead of myself, there are some valid arguments against slashing.

1. Difficulty to pull off on Ethereum due to amount of dependencies

2. If say, Coinbase is censoring and they get slashed, their users are hurt, not them.

29/ Arguments like these and many more deserve debate but I truly believe that this is the best possible solution we currently have for combatting censorship on Ethereum.

And some others might agree:

30/ Regardless, this is a very controversial topic, and it should be decided like anything in social consensus, with rigorous debate.I strongly believe the benefits massively outweigh the costs, but I'd love to hear contrasting opinions.

31/ C. Flashbots Censorship

Within the weeks after the merge a new threat for censorship began to be realized by the broader Ethereum community, and it has to do with Flashbots and the MEV-Boost client.

32/ Before we get into the data, some background:

MEV-Boost is middleware for validators that allow them to outsource the block-building process to relayers, which are able to better extract value from the block.

33/ This software has allowed validators to increase their staking rewards by up to 60%.

But there have been some problems with this.

34/ Some of the relays are regulated and must be OFAC-Compliant. This, combined with the fact that an ever-increasing number of validators have been running MEV-BoostHas lead to 43% of all post-merge blocks being OFAC compliant.

image

35/ It's been especially bad due to the fact that 80% of all validators using MEV-Boost are using Flashbots as their relay, a US-Based, OFAC Compliant company.

image

36/ And as a result, recently a lot of people have been demonizing Flashbots, but it's not fully their fault.

What some don't understand is that although we're not in a perfect situation right now, without something like MEV-Boost, we could be in a worse situation.

37/ Imagine a world where one staking provider, say Coinbase, was way better at maximizing value accrual from the block-building process than any other staking provider.

This would incentivize droves and droves of ETH holders to stake with that one provider.

38/ This would pose a much greater existential risk for the network, as a huge share of Ethereum would be under the control of whoever could provide the best staking returns.

Flashbots, by democratizing their ability to extract MEV, has prevented this.

39/ And although Flashbots attempted some steps to prevent themselves from becoming the most-dominant relay, they still made some mistakes

.I.e. They don't set themselves as the default relay for MEV-Boost, but they also delayed slightly with open-sourcing the Relay code.

40/ Essentially, we're not in a great situation, but we could be in a much worse spot. No, we shouldn't demonize Flashbots, nor should we start slashing validators that use their MEV-Boost software, but some steps need to be taken to bring us to a better state.

41/ Hopefully more competitive non-OFAC compliant Relays are able to be pushed out in the near future, and Flashbots dominance will begin to decrease.

42/ D. The Future, How to Deal With Censorship Going Forward

We've addressed a lot of possible vectors through which censorship could harm Ethereum in this thread, but let's talk about the path forward.

43/ The threat of censorship in it's current form is not terrible, but it can get a lot worse.Right now, if you're trying to send a sanctioned transaction, it will still get through, maybe just with a delay.We need to prevent a future where it won't be able to get through.

44/ We've already addressed possible solutions that could include:1. Implementation of slashing penalties for censoring validators2. Improvement of non-Flashbots Relaysbut there are some other solutions as well.

45/ One possible solution being researched by the Ethereum Foundation is the separation of the Proposer and Validator roles.Under this model, the proposer would just have to accept the block with the highest bid, without knowing the contents of the block.

image

46/ Regardless of the path we take forward, Censorship is something that must be dealt with.Some may have varying opinions, but if the crypto is not truly censorship-resistant, then I believe that the community has lost-track of the mission.

47/ Decentralization and Censorship-Resistance will always be the maximum value proposition of crypto, and we must preserve that.

Through social consensus and technical improvements to the base protocol, we can prevent threats of censorship to our broader networks.

48/ If you enjoyed this thread and would like to see more like it check out the rest of the work being put out by the researchers at @chapterone.

Thank you to @yb_effect for helping review this piece with me.