"A Privacy Layer for Ethereum" by Jordi Baylina // ECC#2 - Buenos Aires 2025

[applause] Hello everybody. Well, I'm Jordi Valina. Currently, I'm very focused in the this project. Zisk is very much about the ZKBM and trying to accelerate uh ZK as much as we can. But today I want to explain kind of a new project or a new idea or something more that's more related to privacy and that is let's imagine how Ethereum should be with privacy in mind.

I mean how we can add a layer of privacy uh to Ethereum how Ethereum should work if we had uh privacy on that. So well it's kind of a dream. Okay, let's dream how it should be. And this is how I think it should be. Let me just see if this works.

Uh I don't know what I need to push. It's the other one. It's this one, right? Oh, yeah. There we go.

Okay. Well, first of all is um all the transactions that are sent to the meool or that are sent to a chain, they should be encrypted. Nobody should see what you are doing. Okay. All these transactions um should be just put in a blobs in a ledger mean just uh commit to these transactions.

Okay, these transactions needs to be processed. Okay, and the state of Ethereum should be also private means that nobody have access to that. So it's a state it's uh encrypted. This is what's represented here is encrypted or RAM state. Okay.

And it should be some let's say some entity um we'll see then how we solve and this is probably the difficult part that should take the encrypted transactions decrypted execute it uh decrypt also the state decrypt execute it and rettrypt that uh state okay and with that so imagine that this entity is a noiscation box it's something that anybody can r this key is a key that this encryption key is a key that nobody knows okay but everybody can execute that okay this is the idea of so fusion fusion will see this is not exist this is not possible today but I hope that someday uh it will work okay but let's focus how this would work I mean how from the user perspective how this would work okay so in order to do that let's imagine for example a normal uh ERC20 uh smart contract the state would be encrypted so the state nobody would be able to see what are the balance but and and only with the constant functions um you can see the balance and the idea is that all the all the permission all all the all the people that can see this balance you can program the ACL you can program these permissions with solidity itself the day is that for example it's in the get balance instead of returning just the balance you first check if the person that's asking for the balance has the right to see the balance so you selective disclosure exactly what you want to to to to see and you can do that in solidity itself. You can also for example the you good things and sometimes maybe you have events. Okay, so you want to notify things to somebody. Okay, so you can have events but the idea is that you can encrypt those events. Maybe those events would be also in the in the ledger and these events could even have attack so that users can recognize some event and can decrypt those events.

So this is very interesting for sending things to one to the other. Okay. So you see priv privacy in this way you see that you don't need to modify too much Ethereum. I mean Ethereum is fully programmable. So all the privacy can be programmed in solidity uh itself.

Maybe it would be some library smart contracts to uh handle all the privacies or the the conversion of the public keys with the private keys ACL. So some helping some tooling that will help to do that. But everything that I said you don't need to modify Ethereum at all. I mean you just program Ethereum uh on you just program privacy using the ABM using Ethereum itself. Okay.

Um so you see that this advantage system is fully compatible with the current Ethereum even it could be updated to that and it would be backwards compatible where everything is public but you can start writing new smart contracts that have uh privacy. um you know the tooling could be embedded in the single smart contracts this goes compatively and um you see that we can create a kind of a standard for privacy to build on top of Ethereum okay and this is one of the advantage of this okay disadvantages well the the disadvantage is that this is not possible today which is a huge but what's the best thing that we can do right now I mean what's the best effort that we can do Well, alternatives that we have first one and uh obvious one I mean trust me bro trust me bro uh way I mean I have a centralized server this is the one that has the state you decrypt this of course you need to trust this central party which is not good but uh well it's it's a good start I mean you can start uh something on that okay but let's do let's try to do it at something a little bit better data central data central server Okay, it's okay. There is another way is okay. Let's try to maybe split the the the state speed the data in a kind of a shared with kind of a shared in a shared secret so that we have like different servers that has this information. Okay.

And then we can have maybe a a stateless processor, okay, that can process this data. Okay. So this has um I mean this model has um I mean has um has a lot of advantages um um but um well it can work of course has it has limitations but also has advantages. this um trustless layer it can be for example a T and the nice thing of this is even if the T is broken some point you can because it's a stateless you can you can replace this T by something else okay and you can keep the TE okay so maybe that separating the that and having a T this should work enough of course T is I mean they are breaking every year and it's it's it's bad at this point okay but it's the best maybe it's the best it's a good it's a tool that we have. Okay.

The other next thing that we can do which is interesting is okay um we split the data we have the TE but the TE can also have a can also generate a zero knowledge proof inside. Okay so this would give us the warranty that we are just uh trusting the T for privacy but we are not trusting the T for the funds. I mean you maybe if the those keys are revealed or you have some problem on the system uh maybe the privacy would be leaked but not the fonts would be leaked. Okay, which is I mean it's bad but it's not that bad that's loss on that. Okay.

Um if you don't want to trust ease we still have some solutions. Okay. So one of those is using goarchs. I mean the idea of co snarks is that is that it's a kind of a multi-party computation but you use a snarks just to uh to be sure that all the parties are acting in the in the right way okay this is I mean the nice thing of this is that you don't need to trust hardware you can do the problem of this is that's quite uh I mean it's quite slow okay um and also um okay and last thing the other way that we can do is um um with uh fullorphic encryption. The best the thing of the the problem of the third one is that you need a lot of interaction with for doing this multi-party computation and then you end up centralizing the servers because all all end up will end up in the same in the same place.

Okay. So the other way is using FHE. The idea of FH is that all these data providers they encrypt with FHE. We have a processor that with FG is just executing the the the state and then the result we use the the the multi-party uh thing to just decrypt uh to the to decrypt the state to decrypt the outputs of this uh of this state. Okay.

Um but uh just [snorts] to mention key I mean the the endo the end game of this is the techn is offuscation. Okay. offuscation is I mean it's not it's not available right now it's definitively possible uh to do it okay it's indistinguishable what's actually possible actually I want to mention here that there is a couple of interesting projects that I mean they are not ready yet but there are interesting efforts to do that one of those is local mixing local mixing looks very promising but it's not I mean it's not proven yet okay but there is some a lot of work that's happening in that direction. The other is machine ID. It's from the Ethereum foundation.

Uh there is a research team that's also working in that and there is other research efforts uh uh in the world that are trying to solve this. But let me say that the the quantity of people that's working in that's researching in offuscation is really really low uh currently. Okay. So we should push that direction. I think that pushing off ofation is important.

And I know that right now looks like impossible. A lot of very smart people try to solve and have not been able. But this definitively is an important piece for having full privacy. Okay. If you want to do an schema like the one that I presented that's very simple.

It would be great to have this of this uh tool. Also fuscation is very useful for lots lots lots of things. I recommend just seeing some of our presentations that I did about the fusation that talking about different uh applications that works very well. So here promoting uh research and trying to find people to look on the fusation is the call that I want to do from here because I think that this technology is super important in regarding the first options just to mention here that we started uh I mean we started recently a team that's working in that we want to write the specification of this layer for Ethereum we want to write a kind of a proof of concept because I'm super interested in the in the usability of this privacy and I think that uh uh start having a smart contracts and start having blockchain applications where they have privacy in mind is a must something that we can advance doing uh uh advance doing right now and yeah that's uh that's very much what I wanted to explain thank you for listening and yeah let's let's do Ethereum private [applause]