Ethereum: Trillion Dollar Security

Transcript

Okay, a few. And of the ones that have read it, is anyone contributing? We got we got we got one at least. Okay, let's go ahead and talk about our security initiative. All right, so Ethereum today, where are we? Uh we have millions of users all across the globe. You know, 500 billion secured by Ethereum. Uh just last month, we had $2.8 trillion of stable coin volume. And all this is being facilitated by a diverse validator set of over a million validators. Um, all secured with 35 million Ether. You know, this is one of the largest public infrastructure projects probably in history. Um, and

it's working seamlessly right now. But the one thing that isn't up here is how much we've lost since 2016 for uh malicious funds being stolen or funds being stolen. And we've lost um more than I think it's 15 billion uh 15 billion since 2016. Uh 2 billion of that is just in the first half of 2025. Um the threat landscape has kind of changed, but we do have some underlying security problems for Ethereum that are going to prevent us from moving to the next uh stage, which is why we have the trillion dollar initiative.

So civilization or Ethereum tomorrow at least to me is uh civilization scale architecture which acts as the backbone of the global economy. You know we're not talking about millions of users. We're talking about billions of users. We're talking about trillions of dollars in values. We're talking about governments and institutions moving on chain and the global economy going 24/7 on Ethereum. And so the trillion dollar uh security initiative is how we're trying to push forward to get us there. you know, we're already seeing institutional adoption. This is happening and it's coming.

So, what is the trillion dollar security initiative? Um, so we launched it earlier this year. Um, the goal is to kind of coordinate with the ecosystem and upgrade the security of Ethereum. Um, and we want to do this in an open source environment where everyone can build on the work of others. You know, not necessarily funding private companies to go make products, but making libraries and repositories that anyone can use, any any team. Um, you know, the goal that we kind of set is we want to have billions of individuals or a billion individuals who are comfortable putting a thousand dollars on chain. So, one trillion. And we also want to have uh institutions and companies or governments that are comfortable storing more than a trillion dollars of value inside of a single

contract or application. And they're also comfortable transacting in similar amounts. Um, that's what we're trying to uh to get to. And in terms of, you know, who's on or what makes up the trillion dollar security, this was started um with co-chairs Josh and Frederick um and I've recently joined. Um we also have ecosystem stewards to kind of keep us on track. Um, we have Sam from the Security Alliance, uh, Zach from Etherealize, and Medi from Sigma Prime, who are, uh, been really great for coordination and communication, um, and kind of helping make sure we, uh, stay the course, so to speak.

Um, and for us, for the choice security initiative to be successful, there's kind of three phases. Um we have this mapping phase where we take ecosystem feedback and we ask you know what is preventing you from growing? What are the security challenges that you're facing? Uh what kind of research are you doing? What's the most important things that we need to look at? Um and going back and forth with these uh organizations. And then once we have a good map of these, then we're going to go to the execution phase, which is where we identify some near-term uh milestones that we can do that'll help accomplish uh and fix some of these goals and also where we can allocate investments for longer term projects.

And the last part of this which is also a very important part is the communication like making sure the ecosystem institutions governments are aware that Ethereum is growing and um becoming more and more secure and is going to be the most secure blockchain for these kinds of things. Um so the first part the mapping phase this is the document that I was talking about earlier that a lot of you haven't read highly recommend it. It's a very good overview. Um, we built this by talking to over 500 individuals, researchers, developers, and client teams institutions protocols stablecoin issuers all across the board.

Um, and going back and forth over a long period of time to kind of distill six key areas that we uh wanted to focus on. Um and for those six key areas that we came the thing that kept on coming up and up was user experience and wallets um smart contracts specifically for security um the infrastructure and the cloud uh reliance that we have and as a lot of people have been talking about the monitoring and incident response um issues with the consensus protocol um those had the least concerns. I think we have a really good track record for the consensus layer. Um but we still have some upcoming challenges u coming

postquantum um changes to the like finality gadget etc. And then we also had um a lot of people talking about the social layer and uh governance and I don't think I have a slide specifically for monitoring incident response but I do want to uh throw a shout out to SEAL. I know everyone's been talking about them but they have done amazing work in this area. um kudos and thanks for all that they're doing. So for the UX and wallet, the main things that uh kept coming up was, you know, key management is hard, like the ability to safely manage your keys and restore. Um approving transactions without not not necessarily understanding what you're doing. Um

granting approvals and permissions to contracts with no expiry and unlimited uh with an unlimited amount of time. And the issue with this is, you know, not only are you giving these permissions when you might not necessarily need to, but it's kind of hard to go and figure out, you know, which contracts and uh which accounts have these permissions and going back and revoking those. Um, a really big interfa a really big problem we've been having is the compromised web interfaces. Um, you know, we're finding that these uh web 2 debt is really killing a lot of the web 3 security uh infrastructure right now. Um, privacy was another big deal. Um, wallets, you know, you can track in the browser. They can see what

kind of tokens you're using. Um, for institutions, this might actually be a legal problem for them as well. Um, and since different wallets uh handle their like core behaviors differently, there's kind of fragmentation as you move around the ecosystem, kind of understand what you're doing. And for smart contract security, the thing that we kept coming up to is vulnerabilities. Um, that's something that we've made tremendous progress on. Um, most of the hacks that we see nowadays aren't actually, you know, due to a vulnerability in the smart contracts. Um, and I think that's, you know, because of all the audits that we're doing. Um, new tooling. I think AI tools are probably going to help out in this in the future. Um, another, uh,

issue that people kept bringing up was the developer experience. um you know low adoption of formal verification, inconsistent testing practices, um frameworks might have uh not defined like defaults uh secure default standards. Um and then risk assessments for deployed contracts as well. So there's not really consistent metrics for evaluating the security of a contract that you're interacting with and its dependencies. And coming back to the institutional side of things, um they might need to have uh like clear lines of accountability for how the smart contract works in order to like uh follow the law. So this is a problem for you know having uh the next phase of the

initiative infrastructure and cloud security is also uh a big issue. Um the RPC and node infrastructure uh being centralized is um a tough problem. Um we have very few RPC providers. A lot of these are hosted on cloud interfaces which kind of centralizes the attack surface for the traditional um attacks. Um provider outages, ISP censorship even things um DNS having a centralized registar which is you know can also be compromised um fishing where you just have like off by one URLs like these like small like annoying things to work with. Um monitor and incident response. Uh the kind of the key issues here were you know if

there's something going on it might be difficult to reach the users or find out or to reach the teams or find out you know how to escalate this or who you should escalate it too. Um there's poor incident response coordination. You know SEAL's been uh taking over for a lot of this and helping uh like large projects adapt which has been really helpful in the community. Um and the last part was just social layer social layer and governance. Um so worried about stake sensor uh stake centralization uh organizational capture of governance um regulatory uh attacks and pressure. Um these are kinds of the six things that we were like most concerned about.

And so we took those things that uh we got from the other client teams and we moved to the execution phase. So here's, you know, six things we can like act on now. They're going to have uh real tangible benefits. Um I'll talk briefly about each of those. So for wallet standards, we have um there's a beta up right now called Wallet Beat. It uh kind of borrows from L2B. And this idea is that you can go through and see all of the wallets and all of the different uh metrics that they implement that are important to you. So, you know, not only do you want to know like, you know, does this wallet have an audit for the current code, but maybe you want to use an open source wallet or maybe you want to use a wallet that, you know, um, you

can use your own node node for. So, you can control for, you know, finding the things that are important to you and then also find secure ones. This also kind of provides a lever for wallet teams to implement the really important issues that users are seeking. Um for smart contract security, we're working with um some clients for producing a large data set uh like a corpus of smart contract vulnerabilities that have been found. Um we're talking to audit providers um and a bunch of different institutions to try to centralize these. And I think this is going to be really valuable for having an open uh data set for AI training. So it kind of removes the moat around um researchers training models

specifically for auditing solidity and this will also enable you know commercial companies that are already doing this to have a better data set for training their own models. Um so transactions assertions this is a hot topic right now it's um still in active research. It's basically this idea of like how can I set a simple set of rules uh for how the state changes for my transaction and if these aren't met then revert. So um one example could be you know I want to swap this token and make sure I get at least this much of this token but also I want to be sure that all my other tokens aren't drained right um and so there's several approaches that we're kind of

researching in the side of this. There's you know the ERC versus the EIP approach. Do we do this on chain in the smart contract level or does this become part of the core protocol for the consensus or in the execution uh protocol? Um still active research for this. Um another very important one is the um verifiable and decentralized frontends. If this is something you're working on, we're doing an RFP phase for this right now on the ESP. So please reach out. Um, you know, we have these smart contracts that are running on decentralized platforms, but these DAP uh frontends still rely on centralized hosting. Um, and their vulnerability to they're vulnerable to uh not only attack but

censorship. And this creates lots of problems in the space. A lot of the uh compromised accounts that we see uh right now is because the front-end um like interface is compromised and the users uh actually interacting with a malicious smart contract where they're signing transactions that aren't in their best interest. Um so if you're working on this, it's a chance for uh an RFP. And then this was the single largest addition uh issue that was kind of identified in the um exe the mapping phase and that is blind signing. This idea of you know I'm signing for maybe I just see some call data, maybe I see

some structured JSON, maybe I just have a hash that I'm verifying on my wallet. I'm not really sure what I'm signing for. Um, and there's been a lot of work put in to fixing this to having a clear signing where you're signing a clear human readable representation of what's going to happen. Um, there's been a lot of work um by the community. The verifier alliance I think has like 8 million contracts for that. They've like done TX decoding for the AIS and verified that the contracts in the source match. Um there's been previous standards done uh 4430 but kind of moved on to 7730. This idea that you can take an external representation of what some

interaction is of a function and map that to a um human readable example. And the idea here being that if we get wallets to implement this 7730, we can have this open- source data set of um a trusted registry of all of these annotations. now becomes really easy for smaller wallet teams to have the clear signing which I think so far we uh all agree is one of the most important things uh for the ecosystem. Um you can also provide that with TX simulation so you can you know see not only uh what you're doing but what the outcome of that might be.

And um the last thing we had really for the execution phase was the wallet for the masses. This is something that's being uh headed by Medi from Sigma Prime. This is like the idea of uh you know your best friend's wallet or your mom's wallet. It's uh a locked down secure on rails system that makes it easy to uh transact and has a better recovery mechanism for nontechnical people. Um some of the key things that we asked for in the mapping phase, you know, we're like sending funds to usernames instead of having to verify these public addresses. And we say verifying public addresses, you know, reading the last four uh digits of the public address in the first four. Um we also want to have uh easy payments

for goods and services. So good third party adoption. Um and things like you know simple swaps uh like basic applications that kind of lock down what you can do and you don't have to worry about getting drained. And kind of the last part about this is the communification phase which is another really important part of this. And um the digital studio team at the EF has been working on this along with uh Josh to kind of map out the things that we're looking at and show what the active research is, what phase we're in. Um and if you want to contribute, you can go to here and see like all right, um this is something that I'm really familiar with and you know, we're still in the research phase. This is how I can contribute. Um really the the most

important thing here that the trillion dollar security initiative is is it's an open-source collaboration for the ecosystem to secure the the ecosystem. You know, not putting it up to like private products in industry, but to come together as an ecosystem to make it safe for everyone. Um so yeah, there's plenty of opportunities uh for collaboration. You know, this isn't about telling anyone what to do. It's about creating a place for like builders to collaborate on the security problems that affect us all all. Um if you fall into any of these categories and you know you you think you want to help and you think you can contribute you know reach out to us. You know this is a very important thing for you know

not only us but the entire community. Um but one key takeaway from this you know it is open source first. This isn't a pitch. Like, don't come and pitch your product and, you know, we're not going to fund your uh your private um private raise. We're not looking for commercial solutions, but community solutions to these community problems. Um do have any questions? That's all that I have, I think. Yeah. [applause] Do we have time for questions?

question. >> Can you please uh show the email again? Sorry. Thank you. >> Oh, sorry. They're bringing it up. >> They're bringing it up. One one second. >> It's trillion dollar security at ethereum.org. We have another question. >> So, it seems like today uh there there's significant problems as you've addressed in the presentation around blind signing. Would you mind just elaborating more on those EIPs that um you were talking about that would potentially fix

them? >> So, I can at least talk a little bit about the 7731. Um I'm not fully uh familiar with all of the previous work that's been done, but this idea is that you have external so off-chain representation of all of the functions and variables in a contract and then for those functions given those inputs you have a human readable uh like interpretation of it. So instead of like calling this function with a source uh source address, a token amount and a destination, it would be you know user X sends you know Y to Alice. Um you know user is delegating this to Bob. Um it's like a text readable representation of

what's happening in code. So you don't have to you know read this like awful JSON uh file of some hexodimal and try to figure out what's going on. Um does that answer your question? I think I got one more in front of you. >> Thanks. Uh, and is the idea then uh with that to somehow integrate it into wallet solutions? Um, one challenge I've I've heard a lot with wallet solutions is that there's often this uh uh tension between wanting to be able to update the wallet and to actually keep the keys safe in a secure enclave that malicious firmware cannot extract,

>> right? Um so in terms of you know while it's not wanting to upgrade um I think we'll remove some of the friction there we're going to provide not only uh libraries for this but the repositories that link these uh implementations and I think the pressure is going to come from like we've seen with walabbe um users are going to want to have this it's going to be an important thing that requires you to you know ascend to the higher levels from that >> sorry I meant hardware wallets >> over hardware wallets. Uh yeah, so that is uh a tough problem like not not only that but even the the interface for which you uh can do that. Um I don't

have the answer to that one unfortunately. I think that's something that the industry is working on though.