Panel | Can Privacy Go Mainstream? | Common S3nse - 2026

So far so good. All right. Um, how's everybody feeling? Good. We're pumped.

I don't like it. Um, again, reminder, Slido, if you have any questions, but I'm happy to take these live questions. It's more, you know, it's maybe easier. um our third panel about privacy. No, I'm just I'm just blown away by all our amazing speakers and I just want to say thank you again for everybody for being here, making the time, sharing your knowledge.

I just it's it always is very humbling as event organizer hurting all these privacy cats. Um but nice to see you all here today. The next panel, can privacy go mainstream with Sebastian Gimu from Midnight Foundation, Kieran from Real Gun, Martin Strubble from Logos, and moderated by Unhappy Ben. Let's go.

Welcome to the stage.

I think we should last person. Okay, thanks for joining us. Um, we've seen, I think yesterday or the day before, like 150 million IDs getting leaked in the US. Uh we've seen flock cameras in the US kind of tracking your every move here in the EU. They just released um or one of the mandates for cars is tracking your eyes to make sure you're always concentrating on the roads.

Um so it's quite a good time I think to talk about can privacy go uh mainstream. Um and the first question I have for the panel is what is privacy for you guys and what does mainstream mean? So I think Sebastian if we start we moved down the line. Um let's go there.

Yeah I mean to me for me personally a lot of privacy uh means sovereignty right. Um so it's a lot of the privacy attempts are basically at times that uh you know infringing on your sovereignty uh and limiting what you're able to do uh by you know basically monitoring and and having the state oftent times be the one to impose behavior through this kind of monitoring either directly or indirectly. And so I think a lot of it in my mind is trying to uh preserve sovereignty for individuals and not have it be infringed by either companies or governments. Um and so what that would mean from a kind of mainstream uh perspective is allowing people to maintain that sovereignty um in a reasonable way. And so uh if the average person is not opting into it, uh I don't think it's because people don't necessarily want their sovereignty.

I think it's because it's, you know, usually too difficult. Um and so for it to go mainstream it has to be that uh you know people are exercising their uh willingness to have sovereignty over their lives.

And uh Kieran I guess the same.

Sure. Yeah. Um I mean you know I think privacy it's yeah it's a very similar vein. I think it's about choice right? Um so the choice to interact with people um you know who and how I want to interact with them.

Um, so like I don't want to be um like forced to be constantly interacting with every single person all the time, which is sort of like how uh like a lot of default is without privacy um online especially, right? And so, uh, privacy is just basically enabling you to have that choice to be like, "No, I I only want to share certain things about myself with, um, you know, certain people that are close with me and, um, like other things with more wider audience that might be a bit more restricted."

And I guess last Martin L.

Oh, yeah. I I don't want to great points. Uh, I I'll use an analogy actually just just to give awarenance to what was voiced here. So I I did my PhD in Denmark and in Denmark it's very common that you have windows without blinds right and everyone can see in your house that's why you get all the good Danish furniture.

Uh but people do it because a they they are not not ashamed right but but they at the same time they believe no one's watching because it's like shameful to be watching. Uh so and there's also like like this this notion of perceived privacy because privacy is is an outcome of that right you either are surveiled or you're you're uh you know having your private space while and and there's also a notion of perceived privacy because that means that you believe that you are no one's watching but actually someone could be watching and I think unfortunately that's a state that we are very much in because the society thinks that you know what we the surveillance that is there is all right or no one is watching um because I have nothing to hide. Uh and uh you know from that I think we we we we need to recognize that there's this kind of routine that we exercise to to to allow others to see what we are doing and control of that. But at the same time it's it's the notion that someone is watching us and and we should note in terms of of the naming that you know surveillance is mainstream of privacy isn't right.

Okay. Thanks guys. Um, I guess you three are from three separate companies. You have Midnight, Rail Gun, and Logos. And each of those three companies have kind of specific ways of being private.

Uh, Midnight uses I'm going to check my notes so I don't up. Uh, Midnight has selective disclosure, rail gun proof of innocence, and Logos has some sort of mixet. I think without trying to be too technical, uh, do you think any of these can go mainstream and how they are, or do you think they already are mainstream? Do you think there are enough people who understand each of your products currently? Because from like a my point of view, I only have used rail gun previously.

Um, and I wouldn't think of that as mainstream. So, I wonder what you guys have an opinion on. It's more of an open question whoever wants to go first, I guess.

Yeah. I think the tricky thing about privacy is that there are many different kinds of privacy enhancing technologies and oftentimes for users, it's not really clear to them which technology exists and why they exist. uh I think even to this day even for a lot of crypto native people the difference between FHE and ZK is one of the most common questions uh we get let alone all the other kind of privacy technology that exists and how they can compare uh and it's often times as uh hard for us at the protocol as well to give people the full range of privacy technologies they might want because you might know a lot of these privacy technologies incur a certain computational cost and so combine different technologies together that each have a different computational cost often times kind of goes beyond ond what people are willing to uh do if not being computationally feasible to do. And so uh you know what we're trying to basically figure out is well how far can we push this and how far can we uh you know enable this kind of composibility these different privacy enhancing technologies. Um and so a lot of our R&D work goes into that topic.

Um but I'd say it's a field with quite a significant number of open problems and I think that's what's made the privacy field in general be somewhat cooperative because everybody's aware that uh there's a lot of research to be done and uh nobody wants to be the one to fully br that uh large R&D cost and so there's been a lot of like really interesting collaborations on this and people uh focusing on on different complimentary approaches

and on please go

uh yeah yeah just to clarify so logos has validated privacy so So we are uh you know co resistant and at the validator level but we also have like a state separation. So we have a private state and a public state execution uh such that you can you can call the the public execution from the private state and remain shielded. Uh but but you know to your point about mainstream I think I think there are like back to my analogy like there are two ways that people actually adopt privacy and the one is when you know you have a freak accident and there's a guy behind your window who's who's looking staring at you 24 hours a day right and the other one is where the social norms change and that's where like you start seeing other people put up the blinds and uh and you realize that actually there's there's a reason behind that and and we believe that you transparency actually has a cost that sometimes it's not made salient like we don't realize there's a cost behind that but if you if you look at like the US equities market you know about 40% of that is uh executed and traded in darkps uh which means that they are executed in private uh so so there's definitely market for private execution and uh we are we are trying to to actually introduce that as as a venue uh and we believe that that we we need to uh unblock this market because like there is a decentralized private market that has not been really uh put out there.

Okay. Just before you go Kieran, so you you say like uh private execution just on your point like from the Midnight Foundation perspective like what product or thing from you guys specifically do you think could go mainstream as a like what you're building now? I think is also on top of that question. Yeah, I mean, so we're building quite a few things, but uh we're focusing on two different uh kinds of core stacks. Uh one is what we have uh kind of in the market right now, which is the ability to do uh private smart contracts um using curve cryptography.

So you're able to uh have uh smart contracts where the inputs of the function are private, but you can uh prove not only the reality execution, but you can also slightly disclose uh parts of that computation. So, for example, you can have a passport and reveal that you're over 18. Uh, but it's kind of goes beyond that where you can even have apps that you use regularly and then later on after the fact partially like uh reveal to a certain person uh part of your history on the application. So we're focusing on a lot of these kind of select disclosure ideas currently but uh looking forward we're spending a lot of time working on latisbased cryptography which is a new field of cryptography that's postquantum uh that has a lot of really promising properties to combine um different kinds of of tools together to build a comprehensive toolbox not just within this like disclosure regime uh but also within a lot of different ideas uh revolving around uh different kinds of privacy intensive technologies like IO and FH that often have some kind of latit space backing as well.

Okay, thanks. And then Karen, I guess.

Sure. Yeah. Um so you know I think privacy right like if we look at going mainstream like using crypto um because I think all of us here are crypto right projects um right like so and and like in the case of rail gun obviously right it's uh onchain um private wallet and then you have um you know like okay it's implementation similar to zcash sapling but you know with x y and z differences and then there's the ability to prove um you know certain statements about that private balance um you know like PPY and all these other things and then you can take that and interact with DeFi. [sighs] there's like there's a lot of technical detail that could be gone into, right? Um, and like I think okay, if you talk about like specifically crypto privacy going mainstream, well, I think the issue is like or crypto in general, we're still anytime that there's some particular um, you know, product or innovation or whatever, we're still having to go down through this like huge list of all the technical details and like that's not mainstream.

Um, I think like sure having access to crypto seems to be mainstream, but like a lot of that is just like, okay, well, I have access to it through someone else who I trust to use crypto for me on my behalf. Um, so yeah, I think I think crypto uh or like crypto privacy could go mainstream. Yes. But um that's also something that would necessarily have to go hand inhand with crypto itself going mainstream. I think they're about on the same level.

Okay, great. Then you guys actually just brought me into the next question. Uh Sebastian, you talk about selective disclosure. Kieran, you mentioned proof of innocence. We've seen the rise of what people are calling compliant privacy.

So that's where you have viewing keys or the ability for I'm not going to call Stark 20 a private protocol or privacy thing, but where you have a kind of an admin key or or things like this. Do you think that this is the only way privacy can go mainstream where there has to be some sort of viewing key or association set or proof of innocence or do you think we can have unconditional privacy like looking forward from the amount of surveillance that's coming to the world in general? Will it only be like selected disclosure privacy or will we have unconditional privacy uh on chain? And I I think I think it's quite a broad question. And so I'm happy to like let any of you answer it, but from my point of view, I think unconditional privacy is something that's very difficult at the moment.

Um, especially with, you know, the tornado cash developers still being in jail and things.

So at Logos, we are trying to do it right because we we have messaging, we have uh decentralized messaging, decentralized storage and decentralized blockchain and we even put uh distributed in an separate app. we are not using a browser distribution because of metadata leakage and other issues. Uh so we are aspiring to do that. At the same time uh you know the product market fit and like where the markets are you see that there's demand for compliance. Uh we uh we absolutely acknowledge that but but we are not you know we are not striving for that as a default for our chain.

uh it is um something that I I believe that if you acknowledge that uh there's this compliance creep where uh you know a little compliance here that that is like embedded in the system will actually propagate through products which uh you don't really foresee uh and that can disintegrate the whole thing. So that that's our approach. uh but at the same time like we we understand that our partners and uh potential uh potential uh business products on our chain will well will need it but but we are not uh providing that as uh as a default uh infrastructure. So we we are living leaving living it basically to them. So it's unconditional privacy for logos and then people can build unconditional privacy for logos and then they can build on if if your users or apps want to build on top this kind of

it's a permissionless place right so uh you know do what you want but I kind of the our principles are u going in the direction of how

good it's pretty cool and I guess rail gun has popppi which

yeah um so so I think like Okay, if you look historically um at these sort of things, right, it's always it's been basically, you know, like all right, you either like go at one end where yeah, private or you have like you give away your view key in which case it's like yeah, you're giving away um you know like perpetual irrevocable view access to to my wallet, right? And so it's just been like yeah well one or the other. Um I think since then right like there's there's been a lot of like uh technical invasion and all and like some of the nice properties about having um some like cryptographically verifiable uh database somewhere right this cool blockchain thing I don't know if you've heard about it um is [laughter] is like it it makes a very very neat source of truth right to go off and like oh we've also had this other cool innovation called like ZK right where ah if you have some you know verifiable input that you can then go and point that and say, "Hey, I'm going to like make some here's the the the state of things that we can agree on, right? And I'm going to go make some some uh statements about that, right? Uh I can go and do that, right?"

Um, and so there's like there's been this like um a march of innovation which allows us to do more and more like granular like okay well giving away my view key might not be like the best the best um solution for like everything but like that was the only thing that I had the choice right um all right instead hey well I'm going to go think about like oh well what's the minimum amount of um information that I need to reveal um to um you know satisfy whatever use case that I have right and you know ideally that's just a binary yes this satisfies the condition um and so I think like PPI kind of nicely fits in that paradigm um for like a use case like rail gun um and I think just you know going forward um as like there's there's more and more like uh different context different use cases and all that that sort of arise um you'll see more of these sort of like more granular popular more um like you know uh more like minimized or like data leakage minimized um methods of um you know satisfying like whatever requirements you have um while still maximizing privacy.

I guess just on that specific point we see the rise of world ID or world from Sam Olman. Do we think like I understand like he says like hey the new world ID is um not he specifically the world foundation say like the new world ID is like minimize it's like secure is it just someone like selling a um something a problem to solve so he creates the problem of like AI in this case and then he sells the world ID to prove that you're a human do you think like from a data minimization perspective from my point of view I don't think I should have to prove prove anything to anyone like you should just accept that like I I really feel like this like do you think that like data minimization is good enough to go mainstream because there's still information that can be leaked and still like draw information from us all to say like hey guys um this is Kieran he is I know ex age and he is proven not to be on some sort of list like that's still enough to know quite a lot about you I guess.

Yeah. [laughter]

Yeah. I mean I mean the thing is that for any privacy protocol you always end up having some kind of uh public state and private state. Um that's basically always the case and it's just a question of um what kind of interfaces do you want to facilitate for uh turning private state into public state. Um and so there's been many ideas around this. Um you know it's like disclosure is one way to think about that but there's many other approaches.

Um and so I think for a lot of people they want some kind of conversion in some way. Uh but what properties if you want for that conversion matters a lot on the project. Uh you know you might want to do like private to public globally so everybody can see the conversion. You might want to do private to public on a person basis. So I disclose to you specifically.

Um how granular is that disclosure? Um I think there's a lot of like uh nuances uh in this. Of course in ideal world we would supply every single possible conversion uh and ways to work with the data because you own it at the end of the day. Um but you know there's only 247 day. So you have to you have to pick something.

Uh and so I think everybody's trying to do their best with uh dealing with the fact that this is a very obvious need for a lot of people beyond just regulatory uh reasons. uh but just being able to effectively handle the data they have um while also trying to be conscious that uh sometimes providing a lot of functionality also you know enables uh bad things to happen.

I guess I think we have maybe five seven minutes left you want to go K.

Sure. Yeah. Um so I I will admit I haven't done a technical deep dive into world specifically. Um and so you know crypto privacy failing to go mainstream yet again because I'm demanding technical [laughter] uh but um I think just like okay looking broadly at the the mechanisms as I understand it there's there's kind of two sort of things there. So like one is they have some like biometric scan um iris scan like secret source which turns it into some um some like hash of your iris scan and then there's like a newer one which is what like it it reuses the public key info that you have on that already exists in like passports and all right um to u prove details that like about that specific identity document.

Um and so like okay if you like examine this and yes okay there might be some use cases where hey I want to prove that I'm a citizen of the Netherlands or um Germany or whatever right um and and like okay well maybe if that's like strictly a requirement of a use case um and yeah this is I guess you know strictly a improvement over having like the selfie of shame um against the white background when it gets leaked, right? [laughter] Um, right. So, so like maybe this is this is like applicable in in those use cases, but also I think um like just blanket saying well here's how it's going to be is not necessarily the best approach because um hey well like something that works in that specific use case might be actually too much for something where you just you just need to like verify that there's you know some uh some like human with a pulse in front of a machine, right? Um and so I I think yes in some use cases but you really need to examine like tailoring it to minimize data.

I think we have seen like eight years of attempts on identity starting from like cycic and other projects and I haven't really seen that product market fit. I think we should build parallel systems and not to source identity in existing structures like nation passports and others. Um so that's what we are trying to do like uh you know I think we are unfortunately heading to to a stage where identity will be blended with the AI but I mean um it's not it's the actions that that are relevant right I I think that we are too much fixated on on on on the notion of like who's the holder of an account which comes from uh from the way the banking system works uh and perhaps we we should really care about the content and uh the actions that uh certain players whether those are artificial or human uh take on chain or or also in other forums and and other places.

I think we have time for one last question. Okay, great. Okay, then uh I didn't write this one down for you guys. Do you think uh privacy has like a product problem and that's why it hasn't gone mainstream yet? like a lot of privacy products, not to pick on rail gun, um it's it's difficult to use.

I think privacy pors, one of my personal favorites, but it's also like not super nice to use. Um do we see a lot of good cryptography, good cryptography and bad product? Um, and do you think that instead of putting all our effort into, you know, going postquantum, which no one knows when it's going to come, should we put more effort into the product side of privacy and building nice products for my mother-in-law to use? Yeah.

Yeah. I mean, I think a lot of the reason that hasn't really happened is because of the the computational cost of a lot of these things. I mean, if you're not in the ZK space, uh you might uh not be aware of how fast things are evolving. I mean, just a few years ago, a lot of these ZK proofs were just computationally infeasible on on devices. Uh nowadays, you can realtime proof of Ethereum uh blocks on somewhat reasonable setups.

Um so you're looking at in a lot of cases over 10x improvement in performance every year. Uh and so you might be wondering if it's going 10x every year, why is are we not getting you know ZK everything? Uh but it's because a lot of the performance gains we get are not being used to optimize just pure performance for performance reasons but rather to add more features. So every time we gain a new performance improvement, we take that new budget and then we allocate that to new functionality. And so that's why you see you know new projects uh you know like logos and midnight that do like private computation on top of data um that uh you know these kinds of concepts are really becoming more and more feasible because of these computational advances.

And so um right now you know there's a quite a large link between technology and uh product because of this fact because a lot of these products uh ideas are only enabled by uh these new papers and these new techniques coming out. Um hopefully as these techniques get a lot better uh there'll be more diversity in the product range um because you'll have you know more budget to basically choose which functionalities you want to spend or which functionalities you want to pick and with the computational budget you have uh but uh right now this is the kind of the state of things. uh but that's why at midnight we also have you know quite a lot of research efforts to try and and push these forward and you know one of the things you mentioned on the qu the quantum space uh one of the difficulties with with postquantum is that they make the size of the proof sign significantly larger. So you're taking a lot of proof sizes that were previously in the hundreds of bytes or kilobytes and turning them to megabyte proofs. Um that's not just a problem because of the data throughput.

It's a problem because a lot of the algorithms you use their performance depends on the proof size and so you're saying that uh you know for postquantum you may have to strip out a bunch of features that we've gained over the year to get this postquantum property. Uh and to avoid this happening is why you know uh we and the Ethereum foundation other people are putting so much money into advancing postquantum. So when this feature becomes a requirement we can introduce this without having to basically strip out uh years of progress on the on the product side. Locos is a movement. So I think I think we are trying you know we are doing a special approach here because we have about almost 50 circles which are like local chapters around the world.

Uh and they are basically like communities that are trying to regain their agency in places where the the n the state failed to provide basic services and they need to organize provide community services. And so they are trying we are using them and leveraging that power to to to define the product. And sometimes we get that uh feedback locally and and you know you learn that one of the most demanded things for example is just a simple registry of local merchants where they can uh leave uh feedback and rating that is unbiased and uh you know censorship resistant uh locally for them. So those are actually things that we can put out in the real world that make impact. Uh and we are trying to start from this ground level while we are also a platform for DeFi and others, right?

But but also we are trying to have this real world impact defined through through these small uh small communities uh and and that actually is like a very grounding force for us that helps us navigate this product question.

Okay. And I think Kieran you can have the last word.

Sure. Yeah. Um I think just generally crypto UX in general sucks. Um and then try to layer privacy on top of that. Um within like technical constraints.

Um I think it's sure like a lot of those have been uh improved. Um and so like okay well the constraints have been lifted and um I I sort of like feel like um you know UX has improved to maybe be like on par with with some other crypto UX. Uh, but we still live in like somewhere where like, oh, to if you use crypto, you must have like the 12 words and like if you're not using 12 words, it's like alarm bells or 24 words, right? It's alarm bells. Um, and so yeah.

Um, sure. But like I think I think like crypto UX in general needs to improve. Keep the answer short.

There we go. I think that's it. Time up. So, thanks a lot, guys. Any last questions?

God damn it. Reluctantly giving the mic to this.

Kieran, don't you get tired of these accusations of rail gun not being mainstream when Andrew Tate is tweeting about it? I apparently don't spend enough time on Twitter, but I also would not consider that mainstream.

Any other great takes from the crypto Twitter sphere? I would say that's success if Andrew I mean that is super mainstream, right? I didn't know he did that.

That's an insane shout out.

Not your favorite K, but you know, just mainstream wise, we're we're out there. Anything else? Anyone else? All right. Round of applause for our panelists and moderator.

Thank you so so much.