Proton’s CTO: No company is going to jail for you

We’ve got the first of a two-part series on the systems that run the world: I’m talking today with Bart Butler, the CTO of Proton, the company that makes private and secure productivity software. There’s a lot of big Decoder themes in this one. That includes how Proton has structured its ownership and architected its products to align its incentives with protecting users. At the same time, Proton faces new challenges and pressures, both at home in Switzerland and from the EU and US government that are putting its values to the test. Read the full interview transcript on The Verge. Links: Proton now offers an entire bundle of office services | The Verge Proton Mail helped FBI unmask anonymous ‘Stop Cop City’ protester | 404Media Proton says it will leave Switzerland if this controversial law Is passed | Vice Age verification is a mess but we’re doing it anyway | The Verge Let’s build a children’s public internet | The Verge Let me see some ID: age verification is spreading across the internet | The Verge Why Chat Control 1.0 is the EU's most Orwellian law yet | Euronews Subscribe to The Verge to access the ad-free version of Decoder! Credits: Decoder is a production of The Verge and part of the Vox Media Podcast Network. Decoder is produced by Kate Cox and Nick Statt and edited by Ursa Wright. Our editorial director is Kevin McShane. The Decoder music is by Breakmaster Cylinder. Learn more about your ad choices. Visit podcastchoices.com/adchoices

• 0:00Speaker 00

  Company spend can be complicated. Managing it shouldn't be. Amex Corporate connects cards, expenses, and payments all in one place. That's less time managing work and more time doing work that matters. Built to simplify, designed to scale. This is Amex Corporate. Get started at go.amex slash Amex Corporate. Terms apply.

• 0:27Speaker 01

  This series is presented by Comcast Business.

• 0:32Speaker 01

  Hello and welcome to Decoder. I'm Neil Appetel, editor in chief of The Verge, and Decoder is my show about big ideas and other problems. Today, we've got the first of a two-part series on the systems that run the world. I'm talking with Bart Butler, the CTO of Proton, a company that makes private and secure productivity software. You probably know it best for Proton Mail, which is encrypted by default, but the company also has docs, sheets, calendar, even a new AI assistant called Lumo, all built and marketed around the idea that they should be vastly more private than the similar products from big tech companies. You'll hear Bart say pretty plainly that the thing Proton sells at a high level isn't really the products themselves, but actually trust. And trust in the software world isn't only about the people who run the companies, but

• 1:16Speaker 01

  also the technology they develop and sell, and the corporate structures in place to make sure that technology is built against the right incentives. This is pure Decoder bait, in other words. The challenge is that Bart also says Proton's mission is very much to succeed at being a viable competitor to big tech, and that means the company has to grow and expand to competitive scale, all while preserving its core values. That philosophy and that challenge are baked directly into Proton's structure and even its physical location. The company and its servers are based primarily in Switzerland, in part because of the Swiss government's geopolitical neutrality. Two years ago, Proton also transitioned to a nonprofit structure governed by a foundation, which is a familiar model used by all kinds of companies that ostensibly operate in

• 2:01Speaker 01

  the public interest, but which has failure modes of its own, as we just saw with OpenAI. Bart and I talked about all of that, of course, but I really wanted to talk to him because he's responsible for the technical construction of some very complex systems that interact with all of those complex politics. I really wanted to know how Bart translates all of those lofty ideals and concepts like user trust into real, privacy-centric products and features that can withstand all of this policy pressure. And of course, there are real examples of this. Earlier this year, the Swiss government came knocking on Proton's door with a request for payment data to help the FBI unmask a protester associated with the Stop Cop City movement in Atlanta, Georgia. Proton complied. They gave the Swiss government that data, which then went to the FBI.

• 2:44Speaker 01

  So of course, I had to ask Bart about all of that and what it means that the US government can use words like terrorism to coerce foreign governments to apply pressure on Proton, and how Proton decides when and how to take on those fights. This pressure manifests in all kinds of ways. Proton is on the record saying it would leave Switzerland and also considered ditching its operations in EU countries like Germany and Norway if the various surveillance laws in those states continue to threaten Proton's privacy mission. Bart told me that these aren't just empty threats and that Proton is in the process of figuring out what it would mean to leave Europe if things get, in his words, more dystopian. There's a lot going on in this episode. We actually ran long because we got so deep into the weeds. We first spent time talking out the broad frameworks that Proton uses before talking

• 3:27Speaker 01

  about the real problems of child safety, age verification, and AI, all of which are testing Proton's values with some of the highest stakes problems on the internet today. But we took the extra time to get there, and I hope you'll think it's worth it. Okay, Bart Butler, the CTO of Proton. Here we go.

• 3:56Speaker 01

  Bart Butler, you're the chief technology officer at Proton. Welcome to Decoder.

• 3:59Speaker 02

  Thank you. Happy to be here.

• 4:01Speaker 01

  I'm excited to talk to you. It feels like Proton sits at the center of an escalating, spiraling debate about how we build technology systems, how we regulate them, and how consumers can protect their data or have any control at all over their data at the center of it. Let's start at the very start. I think most people understand Proton as Proton Mail, but there's now a suite of office products, of productivity products. Describe what Proton is and how you see all the products working together.

• 4:27Speaker 02

  Proton is an ecosystem, if you will, a collection of products that all sort of share the same DNA in the sense that they are, fundamentally are versions of, in many cases, versions of products you can buy elsewhere, but that are privacy preserving. Right? And yes, we started with mail. We also have a VPN. We have Proton Drive, which is our file storage, and photos, and collaborative real-time docs. We have calendar. We have a password manager called Proton Pass. We have Meet, which is a video conferencing software. So we have a whole collection of products that do things, in some cases, very similar

• 5:11Speaker 02

  to other products that you might be more familiar with, but are also privacy preserving.

• 5:15Speaker 01

  One of the reasons I'm excited to talk to you specifically as chief technology officer is the idea that there is a set of products that are familiar, but the company running them is going to behave better than the other company is a familiar pattern in this industry.

• 5:29Speaker 02

  Sure it is.

• 5:30Speaker 01

  And Proton's promise is that the products are actually architected differently.

• 5:35Speaker 02

  Yes.

• 5:36Speaker 01

  Right? That from the very beginning, the way the products are built is actually what makes and keeps the promise of protecting privacy, not a benevolent CEO or benevolent board of directors. We'll come to that. There's some of that in the mix here.

• 5:48Speaker 02

  Yes, there's some of that, too. There's some corporate structure stuff, but there are actually, I would say, two primary structural, maybe, or systems, you could call them systems engineering, right, at a broader scale, but structural constraints on how Proton operates. The first is that we encrypt all the data we can. Right? So if we wanted to turn around and sell that data to somebody, we can't. It's mathematically not possible for us to do it. Right? This also has other benefits, right? We can't easily lose it to hackers or other interested parties. And we, you know, there's some data that we can respond to for, say, legal requests, but, you know, there's some data we can't, right?

• 6:33Speaker 02

  And this helps us, basically, we can't, we can't give up data that we don't have access to. The second is the business model. It's not, I mean, there are other SaaS players, of course, that do this. I don't think there are a whole lot of B2C consumer-based SaaS players at our size who do this, but our revenue model is getting paid by our users. Right? So we don't sell ads. The products are not carrots to get people to come in and give us our data so we can sell it to advertisers, right? And that means that those users, the ones who pay our bills, pay our salaries, allow

• 7:13Speaker 02

  us to grow the business, if we were to betray them, then that would essentially undermine the value of the business. Right? So we have tied the value of the business and the growth of the business to protecting our users so that our interests are aligned. And this is also very important because, you know, temptations are a thing, right? People respond to incentives, and we have structurally arranged our company such that those incentives are aligned with the people to whom we have promised to protect.

• 7:40Speaker 01

  If I was to very reductively summarize what you just said is, we take money from consumers and when we sell them is encrypted versions of popular services that they rely on, like email, like a VPN, like an office suite. Do you think consumers understand that what they're buying is the technology solution? Or are they buying the promise of privacy? Or is it some mix? Because I'm not actually sure consumers really understand at mass scale how it all works. Right? They think their iPhones are listening to them.

• 8:10Speaker 02

  I will tell product people and engineers, you know, in meetings about new features, if you've, if you've mentioned the word encryption to the user, you've already failed. Right? I mean, people don't understand what this is. That's fine. Our goal is to make products that are as usable and as functional as our competitors, or more, or more functional. I mean, I don't want to sound entirely derivative, right? We do have, we do have features that nobody else has that are often geared towards privacy and functionality for people who need confidential communications. However, in general, we're selling, we're selling the promise. We're selling the promise that we're a different kind of company. We're selling the trust. Right?

• 8:54Speaker 02

  And that trust is critical. Without the trust, that is where the real value of the company is. Now, that trust is backed up by technology. Right? But, but we're selling the trust.

• 9:04Speaker 01

  The reason I'm pushing on it, and specifically, I'm happy that you mentioned trust, is the big tech players will all make the same kinds of promises about your data being private. Facebook will happily tell you that they don't sell an ounce of your data. It's actually not in their interest to sell the data because the ad targeting that they do depends on the data being theirs and not anyone else's. We can get into this for days and days and days. I'm just curious where you see that trust being expressed, or where you feel like that trust is most communicated from Proton? Because if you ask me as a more technical person, I do look at the architecture of, it's all encrypted and probably mathematically, it's impossible to get into. And sure, we'll come to how much metadata can be shared with authorities because there

• 9:49Speaker 01

  is some debate about that. But that's the piece that resonates for me, as opposed to, I have to trust you or a board of directors. There are some other people who just look at the promises and take them at face value. Then there's some set of regulators that say, actually, we have a bunch of other interests in being able to see the data that goes on here, and we actually don't trust you to be a good player in the ecosystem. So when you think about trust, is it, does it come down to the data is encrypted and that's the core promise we're making, and anything that breaks that breaks the whole product? Or is there some other dimension of trust that is important at Proton as you build the systems?

• 10:29Speaker 02

  So and encryption is obviously important. It's the gold standard, and it's what we strive to. However, there are definitely features that, you know, we...

• 10:40Speaker 02

  We privacy, at least to me personally, is about control, right? And there are some times where I want an integration with an external service or something like this where I'm going to have to. I mean, my choices are no integration or my choices are breaking into encryption. And our goal is to make this, you know, the user has an informed choice and control over who sees their data. So it's not the same as never share your data and, you know, never share your photos with anybody. But the point is, instead of sharing your photos with somebody like Facebook, who might, you know, monitor, it's sharing your photos with, you know, grandma and grandpa, and only grandma and grandpa, right? So, uh, so that's not to say that, uh, the, the, the encryption is everything.

• 11:23Speaker 02

  There's the fundamental engineering, which backs up the trust. Like people like you look at the architecture and say, okay, this is encrypted. This is important to me. And then you go tell somebody else, you say, I trust Proton because of this, this and this. The other person, all they know is, hey, I like, I know Neelai, I trust him. And he says Proton is good. There's a whole cohort, much, much greater than the techie cohort. The techie cohort is our core, of course, and has been since the beginning. But there's a whole cohort of people who trust Proton because other people they know, whether, and by know, I mean, it might just be people who they trust on Reddit, right? But other people they know say Proton is, is there.

• 12:03Speaker 02

  And we've all been getting object lessons in, in the, you know, rules, bylaws, laws, other things are, uh, can be worth something, but personnel really matters, right? Who enforces them really matters. So we have the technical layer, which is designed to constrain what we can do technically. And then we also have the, you know, the legal, the corporate structure layer, which is, it's defense in depth, right? All of these things are interlocking guarantees to our users that we are trustworthy. And ultimately, that's the important, the most important thing about the business.

• 12:35Speaker 01

  I do want to come to the corporate structure. It is decoder after all, but, uh, just one more turn on the product set itself. You started with Proton Mail, you've got the other suite of products, including now an AI assistant. You talked about the fact that the business model is the consumers paying money directly for the products. Is Proton Mail still the core product that's making the most money? Are the other lines of business growing? How does this look?

• 12:56Speaker 02

  We don't disclose direct financials. Mail and VPN are the two oldest products. And, uh, as you, you might expect, they are the two largest products still. But, uh, you know, we also have a lot of people who buy bundles, like, you know, and buy multiple products. We try to make the products work well together in an ecosystem. Some products are more tightly bound than others. And the new products, um, the more recent ones, calendar, calendar is very tightly tied with mail, of course, but we also have, uh, Drive and Pass. Those are not as big as the older products. They have less of a, less of a head start, if you will, but they are growing rapidly. So, so they all contribute, but yeah, they contribute differently.

• 13:36Speaker 02

  Uh, often, I'm not going to say it's exactly how old they are, but, you know, those with, those that have had the bigger runway are usually bigger.

• 13:45Speaker 01

  One of the patterns with consumer productivity software is the, the suites get bigger, the bundles get bigger. And then the companies realize the class of customers that will actually pay for increased productivity is enterprise. And we kind of see this over and over again in the space. Notion did it, Dropbox did it, ever. Like I can go on and on and on of companies that we covered as consumer software companies that eventually pivoted to being enterprise companies. Does Proton feel that same pressure that you're going to have to go have more corporate clients, enterprise clients to grow, or are you still focused on consumer?

• 14:17Speaker 02

  I think we definitely feel the pressure. There's, uh, what is it? I don't know if it's apocryphal or not, but the old John Dillinger quote, you know, like, why do I rob banks? Well, it's where the money is, right? There's a little bit of why do you go B to B? Well, it's, you know, it's where the people who have budgets and are willing to pay for this, this goes. I think, um, I think we're definitely considering, and we have, even, even as a primarily consumer focused business, we have a lot of people who, small businesses in particular, who use our products for business purposes, right? You just use the consumer. And, and we've, we've also made forays into, uh, small business offerings and things like that. I think this is something that, uh, that growth will probably demand.

• 14:59Speaker 02

  And I think there are a lot of businesses for whom things that we offer, confidentiality, uh, um, you know, non-US based, uh, you know, European sovereignty type stuff in particular. Um, and, and just our reputation that are very appealing to those businesses, right? That said, we have to do a balance, right? And today, our business is B to C. We're not going to jettison the B to C business. I also think that, uh, having the B to C business being as strong as it is, is, is a, is a competitive strength. There aren't, I mean, as you said, there are several that have transitioned to B to B, sort of from B to C.

• 15:40Speaker 02

  But at the same time, I think, uh, there are a lot, a lot more B to B players that start it, start off B to B and stay there, right? And, uh, one of the, one of the things that, uh, that, that we're looking at, you know, communication between businesses and their customers is something that is, uh, is certainly important and certainly confidential, confidentiality is important there. And you see things like with, uh, you know, WhatsApps, um, forays into, you know, having businesses connect to people who have consumers who have WhatsApp. Like, I think that when the time comes to really make a push to B to B, the B to C, uh, user base

• 16:22Speaker 02

  and product suite will be a, will be a benefit to us.

• 16:26Speaker 01

  One of the tensions there is AI sort of infiltrates more and more businesses is the frontier model companies want every ounce of data. I think a bunch of big enterprises are very leery of giving a bunch of data to frontier model companies. The AI works best when they have a bunch of data. And so there's this big choice about how much, how much of your business will you expose to Claude? How much of yourself will you expose to Claude to get the most value out of those models? Proton sits right in the middle of that with a technical architecture that you're saying would provide choice. But it also seems like the, the game for a lot of these companies is to just hand everything over and say, go run my business, AI. How, how do you see that working with your enterprise customers today?

• 17:07Speaker 02

  It's a fraught decision, right? It's giving all their, all their sensitive data over. And they feel like in many cases, they don't have a choice. And, uh, in some ways, that's why we developed Lumo, which is our AI offering, which I, I think may have left off the list before, unfortunately. But, uh, and we, I think we just launched Lumo 2.0, which is a big revamp of, of the models we use. And part of that is, you know, the goal of Lumo project in general is to have something which integrates with the rest of our products and can do this in a safe way. This is to address this sort of trade off between, do I give random AI companies, possibly in different countries, possibly, you know, have compliance problems, all those other, all those other things? Do I give them all my sensitive business data? Or can I do it in a way that is still, is still more protected?

• 17:52Speaker 02

  Now, it, you know, in many ways, uh, there are still trade offs on our side to make, but we keep it in-house, within Proton, with the guarantees that we give. Uh, that should be a more attractive option for those kinds of workloads. And that's what we're, that's what we're hoping as we develop Lumo.

• 18:08Speaker 01

  I think this does bring us to the, the decoder questions about structure and, and organization. Cause that's where it feels like the structure has to be where the trust lies, not necessarily the, the technical architecture. So how is Proton structured today? How many people is it?

• 18:24Speaker 02

  Proton today is, uh, approximately 650 people total. That also, uh, that includes engineering, support functions, marketing, et cetera. It also includes customer support, which we, which we do in-house, we always have. Um, we are a corporation, Proton AG, a Swiss corporation. However, a controlling stake in, in Proton AG is held by the Proton Foundation, which was seeded with shares from Andy, our CEO, and other early employees and has a controlling stake. There are other fellow travelers with, with, I would say, a similar, uh, in somewhat similar structure, Signal and Mozilla, and our reasons are the same. We have a, um, you know, the Proton Foundation controls, has a controlling stake in the Proton company

• 19:09Speaker 02

  and has, as, and is empowered to protect the mission. It's a Swiss foundation, which is, uh, I've been told, I'm not a lawyer, but I've been told is very difficult to change. And its job is sort of the guardian of the values and the mission of Proton. So if, if Proton, the company were to stray from that mission, the Proton Foundation would be empowered to, uh, correct, to correct course, if that makes sense.

• 19:30Speaker 01

  Has that ever happened?

• 19:32Speaker 02

  Uh, those, no, thus far, no. We have, uh, our founder CEO, Andy, um, he was, Andy Yen was, he founded the company, as, as founder CEO implies. And as long as he is, uh, in charge, I don't think that'll be necessary. But, you know, if he gets hit by the bus, there's, there's, there's still a, uh, there's a corporate structure which is designed to protect the mission going forward and to make sure the company doesn't stray regardless. And, and because of this ownership model too, the company is insulated from, say, some sort of takeover that, or, or purchase that could also redirect its, um, its, uh, its priorities. But I do, I, I, I do say that, I mean, I think, I think that, and the tech,

• 20:15Speaker 02

  and, and the business model basically are interlocking protections against, against us, you know, betraying the compact that we have with our users.

• 20:25Speaker 01

  One of the things that strikes me as I talk to more and more companies that have transitioned to this kind of foundation model, is that that structure is essentially there to insulate you from the rapacious demands of capitalism, right? Like you, you aren't being pressured to make as much money every quarter as possible, to do the things that would lead you to make the most money. And that means maybe the technology can develop in a pure and idealistic state of protecting privacy instead of chasing the dollars. Do you feel that insulation, right? I mean, you do have to make money and pay your employees and, and grow in some way. But what's the dynamic for you architecting the products?

• 21:01Speaker 02

  The short answer is no. And, um, the reason for that is that we have to complete, compete in capitalism. Our, our main competitor is big tech, right? Even though we are much smaller than big tech, you know, where our users come from, when we convert people, when we get new, new customers, they're big tech in general, right? And, um, We have to play the same game and I, our corporate motto, this might be a little cheesy, but our corporate motto is, you know, privacy by default. And that default part is doing a lot of heavy lifting. I mentioned before that our goal is to design products which are easy to use and secure. Right? There are plenty of tools that are secure and nobody but a few, you know, a few experts use them.

• 21:47Speaker 02

  And that's fine, like I'm not criticizing the existence of those tools, but our goal is to make tools that everybody can use without understanding the cryptography, without even knowing that it's cryptography that they can use and that are as easy to use and as feature filled as are unencrypted, you know, unencrypted and essentially data mining competitors. Right? It's a tall order. I'm not saying we're, I'm not saying we're a hundred percent there yet, but that is the goal. But the default word in that privacy by default means that we have to be at the scale where we can offer a real alternative big tech and small startups, small scale ups, you know, being a hundred times or 10 times smaller than, than big tech is not going to cut it. Right? In order to do that, we need to grow.

• 22:33Speaker 02

  So growth is actually part of the mission, if you will. And that means that we have to be, the word not rapacious, but we have to be as hungry and as efficient and as growth minded as we possibly can, because that's part of the mission to grow big enough to actually challenge the paradigm. We can talk about all kinds of ways capitalism is kind of broken right now, but like we have to play the game.

• 23:00Speaker 01

  All right. I'm going to make a comparison that you are going to hate.

• 23:02Speaker 02

  Okay.

• 23:03Speaker 01

  I'm just letting you know, I know you're going to hate it. Maybe the most famous, we'll build a foundation to protect ourselves from the demands of the market and make sure the thing is healthy is OpenAI, which basically killed that structure in order to chase an IPO, right? Like, and maybe they still have really important and idealistic ideas about how Adrian should be developed. Well, maybe, maybe there's some people at that company who really feel that way. And it just didn't work, right? They needed to chase growth in very specific ways for whatever reasons that they felt. Maybe it was money, maybe it was just in order to take on Google search, the way that OpenAI felt like it wanted to take on Google search. They had to change the structure of the company. That obviously happened, right? It's like one of the most like apocalyptic foundation moments that has ever happened.

• 23:47Speaker 01

  Like this thing just like exploded or imploded onto itself. When you look at that and then you look at, okay, we have to grow in order to be the default. We have to grow to be big, but we're making this promise that Google doesn't have to make or Microsoft doesn't have to make. Where is the tension in that? How does that express itself as you design the architecture of the products, which might preclude some opportunities?

• 24:11Speaker 02

  It's not, but what could also be our corporate motto is go big or go home. Right? We don't set modest goals in that regard. But, but, but I think that there are, there are a couple of things that, that, that make it different. I mentioned before, a lot of constraints, but one thing I didn't mention is we don't have VC

• 24:30Speaker 01

  Yep.

• 24:30Speaker 02

  investors. We don't have private equity investors. We aren't burning other people's money with VCs burning down, you know, breathing down our necks that we must exit soon or otherwise we go belly up. Right? We, we, our goal, we built a sustainable business. We reinvest the profits from that business back into the business. And this insulates us from some of the pressure, which I'm sure OpenAI felt, given that they were, you know, lighting enormous stacks of money on fire all the time. Right? But that said, you know, personnel is policy, right? There's certainly always this risk. And at the, at the same time, you know, sometimes you find that we, we, we've found out a lot recently that sometimes, you know, rules or otherwise or norms or guidelines, unwritten or not, they're not worth the, the paper that they're written on. But that's also where the architecture comes in.

• 25:15Speaker 02

  We make choices about the tech stuff. And obviously I'm on the tech side of this business, but we have made choices. And, and some of that is for business competitive reasons, right? We want to sell products where we say, we can't access your data. We don't have access to the data. We can't lose it. We can't sell it. We can't, you know, we can't do this. But that also constrains us from deciding one day to turn around and sell it. Because we have locked it in a box, right? And we can't access it and therefore we can't turn around and say, you know what? Sorry, guys, we changed our mind. We're actually gonna, gonna mine all this data and go. Right? Now, is anything perfect?

• 25:55Speaker 02

  No, but I mean, the structure is designed such that, such that we have these interlocking, as I said, controls. The proton foundation is barred from selling proton, proton A G is as far as I'm aware. Right? But even if some reason, some happens and this happens, you know, the value of proton is in the reputation. And then we said this, this has been a sort of a deliberate choice. The value of proton is in the trust that we have. You know, if Google bought us, it would have no value because Google does not have the credibility to run proton.

• 26:27Speaker 01

  That's just the default of Google buying anything. I just want to be very clear about it.

• 26:30Speaker 02

  I know, no, they shut it down and then it's gone, right?

• 26:32Speaker 01

  The clock would start ticking that day.

• 26:34Speaker 02

  Yeah, I know. I, I know. This has happened before.

• 26:38Speaker 02

  I, I'm sure, but that is, but the proton foundation structure is designed to say that, no, the company cannot be sold to a buyer. Right?

• 26:47Speaker 01

  That's the macro structure. Just tell me about the more tactical structure inside the company and spoken in product. How is proton itself structured or is there a team that makes the Lumo system? Is there a team that makes email? Is it divisions? Is it functional? How's that work?

• 26:59Speaker 02

  We do have a division structure, a business unit structure. We have certain products are bundled together like mail and calendar because they're very, because they're, so they're bundled into the same division. Other ones are separate. So, you know, we have Lumo, we have Drive, we have Pass, and we have VPN. These are, these are separate, separate divisions. You make choices about corporate structure based on the, what communication you want to be easiest. Right? So that's designed to allow people to move fast and to have autonomy and, and, and make decisions within their product. And then we also have teams that work sort of cross organization teams or support teams and things like that. And on that side, we off, the trade off of the business unit structure is there.

• 27:43Speaker 02

  We sometimes have to work harder, right? It's a little bit like corralling cats, right? Where, where you have to make sure that, okay, I'm building this feature that touches all the products. I need to make sure that all the products are on board and they have it in their planning so that we can ship it correctly. So we've made a deliberate decision to sort of prioritize in product communication. And then we have to work and then we try to compensate for the shortcomings of that in our cross product communication. And then as we transition to more and more ecosystem based stuff, we may make, make different choices about the, about the corporate structure to support that. You know, the whole Conway's law thing, right? You ship your org chart, you know, but sometimes you, you, you want the best of both worlds.

• 28:26Speaker 02

  And this is where process comes, comes, comes along. It's not always sexy, but it's, you find it super important as you scale organizations for sure.

• 28:34Speaker 01

  You might be the first decoder guest to, to make the connection directly between the fact that I ask everyone how their company is structured and you ship your org chart, which is why I always ask, because

• 28:44Speaker 02

  Yeah.

• 28:44Speaker 01

  it's the fundamental truth.

• 28:46Speaker 02

  You do, and we, and we struggle with that to some degree. We try to compensate for it, but it, yeah, you do ship your, your, your chart. And therefore, if you want your products to look different, you should adjust your org chart to support that difference, right? I mean, ultimately.

• 29:00Speaker 01

  The other decoder question I ask everybody is about decisions. You have a lot of decisions to make. You are trying to build a new kind of product architecture against a lot of the same constraints as your competitors. How do you make decisions? What's your framework?

• 29:12Speaker 02

  We are a, you know, founder led company, right? Andy, Andy started proton.

• 29:22Speaker 02

  I was, I was, I think employee six, I want to say, or something. That, that's a funny story in it, in itself. You know how I met Andy?

• 29:30Speaker 01

  How's that?

• 29:32Speaker 02

  He was my grad student at the time. I was, I was a postdoc at Harvard and he was my grad student. And then he comes and finds me in Silicon Valley after I left physics. I was like, hey, you want to join my startup? You can be CEO or CTO rather. But yeah, so he's still heavily involved. He, you know, he has a lot of input on, on product direction, strategy, and he's a visionary. Right? And I think he's responsible for a lot of, a lot of proton success. He's willing to take risks, this kind of stuff. And I obviously believe in his, his, his leadership. I wouldn't still be here after 11 years. That said, I think one thing that sometimes happens with visionaries is they need to surround themselves with people who will challenge them when they get maybe a little too far over their skis

• 30:14Speaker 02

  or, or to also make sure that we have ideas from other places. One thing that we, that we really try to make sure is that, you know, ideas are judged on their merits, not who, not, not, not their origin necessarily, right? We have a senior leadership team, which, whose job is to, yes, execute, but also, you know, bring new ideas and sell them and things like this. We try, I'm not going to say that we always, that we're perfect in this. We have plenty of things that we can improve, but you know, we try to push decision making down the orgs. The whole point of an organization is that you have, it's a way to align lots of different people in the same direction, right? And the trick is always how do you get alignment in the same general direction

• 30:56Speaker 02

  while still having autonomy so that you're not micromanaging everything? And I'm not saying we always get the balance right, but that's the goal here. And when, you know, we, we do this sort of orientation when we hire new people into proton. And one of the things, one of the things we always say there is like, I want to hear, you know, you, you're, you're not used to our, our way of doing things yet. So I want you to look at our way of doing things. And if you've experienced another employer or you think this is either crazy or inefficient, I want you to tell us, right? And we were very clear, yes, of course we have a hierarchy, but it's, you know, level, and part of this is our size, of course, but we very rarely have more than, say, three or four levels of management. And we always say, look, if you need to ping me, you need to ping the CEO,

• 31:41Speaker 02

  you need to directly talk to people, just do it. It's, it's another thing. And again, I'm speaking for myself, but I think this is sort of a comment throughout proton is that I tell people, I, I'm not promising or obligated to agree with you, but you're never going to regret telling me what you think. We have, and this is, I credit Andy with this. We have very little internal politics. Maybe this is something just because we're not very big yet, but we have very zero tolerance for, you know, fiefdoms or internal politics. And that also makes, makes things that there's, there's very much a sense that we're all pointing in the same direction and that it's not my, not my division that I'm trying to grow. It's proton in general. Some of that comes with being a mission driven company.

• 32:24Speaker 02

  I don't know if you've ever had this experience, but for those of you who haven't, I, I've had both, right? I, I was. I was a physicist as part of the CERN collaboration a long time ago. And there you have, you know, fundamental nature of the universe. I consider that kind of a mission driven occupation. I also had a stint in Silicon Valley afterward, which was, uh, which is fun and interesting technical problems, but I, I, I, I wouldn't, you know, I missed that, right? And with Proton, the mission is really, uh, we have to be careful that it doesn't paper over organizational problems that we should really solve. Let's put it that way. It's, it's, but it really does make the job easier when everybody is sort of aligned in that regard.

• 33:10Speaker 01

  We need to take a quick break. We'll be right back.

• 33:22Speaker 00

  Real innovation is powered by real people. This is Amex Corporate. Trusted experts, smarter technology, 24/7 dedicated support from day one to answer questions, help set up policies and approvals, and consult on spend capacity as your business needs evolve. Plus, on-demand resources help level up operations at every stage. That's the trusted service you've come to expect. Built to simplify, designed to scale. This is Amex Corporate.

• 33:58Speaker 01

  We're back with Proton's Bart Butler. Before the break, we went over the decoder questions, but now I wanted to put Bart's answers into practice and discuss how Proton is maneuvering an increasingly threatening policy landscape, both here in the United States and in Europe.

• 34:11Speaker 01

  Early on my career, I had no idea how to manage anyone. And I went to a bunch of people and asked them a bunch of questions. And one of the smartest mentors in all of this told me very seriously, she sat me down and she's like, look, you don't hire people to make them like you. You hire people to change your organization. And I've taken that to our, and maybe at 15 years into this, I'm like, that's actually a pendulum, right? You, you need people to buy into what you're doing. Otherwise, every new person you hire is going to radically disrupt what everyone else is doing because they're maybe over empowered and they're, they're not actually on the same page. Proton has a mission, right? You're describing as a mission driven company. How do you strike that balance of you want to hire new people and get the outside perspective on what are we doing wrong?

• 34:55Speaker 01

  And then not actually get knocked totally off track? Because it seems very important inside of a company like Proton.

• 35:01Speaker 02

  I mean, culture is extremely important. It's extremely important. And that comes from the top as well. Andy, if, if, if you were to ask him, what's the most important thing in any business, he very likely to say, to say culture. And as a result, we, you know, we try, we're very careful about who we hire. We hire relatively slow. I think we could, I wish we could hire faster and maybe, you know, I think hiring is one of the things that we could be better at, but we don't do these, you know, massive hires, massive layoffs, things like this because, and, and we don't hire so fast that we dilute the culture. Like we want, we want to integrate people into the Proton culture, not necessarily. And yes, sometimes we have to change it.

• 35:44Speaker 02

  Sometimes we have to evolve it, but we want that, we want that to be done in a, in a, in a deliberate way. Um, I think, yeah, I also had a, maybe a similar evolution. I've been there through most of, all of Proton's growth phases. So, you know, early on I wrote a lot of code, then I was, you know, effectively a team lead, then I was a manager of managers. At some point I was running all of Proton's engineering, then I, then I was more, you know, handed off some of the management things, but you know, kept the sort of CTO, technical direction stuff. And in that course, uh, definitely made a lot of mistakes, mistakes too, right?

• 36:26Speaker 02

  Early on, it was a, I had to, um, you know, teach myself not to make other people like, like me, not to, maybe not micromanage, but not to tell, not, not to just tell people what to do and have them execute it. And then I went through a phase later where I let people, you know, I, okay, people need to learn, people need to make their own mistakes, people need, you know, I want to import people with expertise, so I, I, I'm not infallible. Let's, let's, let's do this. And that was probably too permissive and it caused, you know, we had some, nothing catastrophic, but we had some expensive mistakes that I had a bad feeling about, but I let go through anyway, because you know, because I was trying to do this.

• 37:10Speaker 02

  And so I think, you know, moderation doesn't tend to be the sexiest thing to sell, but it is, it's, it's a balance, you know? You don't want to mandate how things are done, but you also want to make sure that you're there to stop people from doing like truly catastrophic or expensive decisions that you, you can see the brick wall in the, you know, in, in the distance and you want to make sure that doesn't happen. It's not the most dramatic answer, but a lot of this is finding the right balance there. And it's a certain amount of moderation.

• 37:40Speaker 01

  The reason I spend so much time on the technical side, the corporate structure side and in the culture is because Proton faces a lot of pressure and all of this, as you've described, is designed to resist that pressure, is designed to build products that can't be broken by that pressure in different ways. Let's just start with, I don't know, the governments of the world, which are a lot of pressure and have found lots and lots of ways to get past the kinds of controls you put in place to protect user data. We'll just start with the splashiest one. In March, a report from 404 Media found that Proton handed over the payment data of an account called Stop Cop City, which is located in the United States. They handed that data to the Swiss authorities, who then gave that data to the FBI and that led to their identification. And this is metadata.

• 38:25Speaker 01

  I don't think it's actually the, the data, the, like the contents of the emails. Proton's argument is, look, we never actually gave anything to the FBI. We just complied with a legal request from the Swiss government. Let's start at the very basics. What is the Swiss government legally allowed to request from you? And it, does the fact that they can just serve as a proxy for the United States government undermine any of this trust or put any novel kind of pressure on your structures?

• 38:47Speaker 02

  Any company, anywhere is going to have a jurisdiction, be subject to jurisdiction. There is no country, there's no company or an individual which is above the law, right? And nobody, um, you know, no company is going to go to jail for you, right? That said, you can arrange structures such that there are safeguards here. And our, our safeguard, for instance, is that we are a Swiss company. And we've actually been asked repeatedly, hey, can you just respond to requests from friendly government agencies? And we, we have repeatedly said, no, because it can't be our job to decide what is legitimate

• 39:32Speaker 02

  and what is not. That is not something that we can take on, right? So what we do is we, um, we engineer our products to have, you know, within constraints like making a product that people want to use and can use and do the job, right? But we engineer our products to be as private as possible. And the Swiss govern, and, and we are subject to Swiss jurisdiction.

• 40:00Speaker 02

  Mutual legal assistance treaty request, MLAT requests come in from governments, the Swiss, uh, Swiss authorities decide what is legitimate, what is not, we have no stake in that. And then they issue an order and we, and we comply with those orders. And that's the way, that's the way it has to be. And we, and we very deliberately chose our jurisdiction in a way that, you know, the Swiss are famously neutral and they also, you know, have a certain, like every government's made up of humans, but they have a reputation for, for, for being reasonable people. And as a result, and that system has worked pretty well. In gen, in general, there are countries that are less trustworthy than other, less trustworthy than others. And those requests, we don't have a lot of visibility into where the requests are coming from, but those requests we have on good authority are usually not honored.

• 40:45Speaker 02

  Whereas, you know, this, people's opinions may vary these days, but US, you know, the FBI requests, they tend to be, they tend to be given a certain presumption, but they're still a presumption of legitimacy, but they're still evaluated by the Swiss authorities. And then what we do is we comply. We have no, we have no, uh, discretion here. And this would be the case, the case for any of them, but we have arranged the system such that we think it is the safest, one of the safest that can be constructed, you know, and, and, and, and stay legal.

• 41:17Speaker 01

  So let me just ask you about that, because you are a systems person and you're describing a system. The failure point in that system, as you're describing to me, is the Swiss government is going to make a bunch of decisions about what you have to comply with. And the United States government, in this case specifically, I, I think has realized if they just say that everything is terrorism, the mechanisms for data sharing, sort of the floodgates open, right? So in this case, this is an account called Stop Cop City. They, they said this is terrorism. Here in the United States, whether or not the government's claims of everything being terrorism are legitimate or not, I think are wide open for debate.

• 41:51Speaker 02

  Sure.

• 41:51Speaker 01

  It feels like they found what you would describe as an attack vector on the Swiss government's mechanisms, where if you say these magic words, the Swiss government will come to you and start asking for metadata. Does that feel appropriate? Does that feel survivable?

• 42:04Speaker 02

  I think a lot of this stuff, it's at some point, it's, it's going to be up to people, right? We, we've done the best we can and I don't think, you know, it's never going, people are going to have different opinions about what, which requests are legitimate or not, but we've done the best we can in saying that, okay, the Swiss authorities will decide, we will comply with whatever they say because we are, we are a Swiss jurisdiction. In the meantime, we do and can and, and arrange our, and this, of course, there are laws that govern this, but we do whatever we can to minimize the amount of data that we can give, right? I mean, payment data is sort of, there's, there's not, we can't encrypt it, then we use payment processors, like they can get it, they can get it from lots of different sources. So if you have a, if you have a credit card attached and, and there's a legal request

• 42:48Speaker 02

  coming in, there's not a whole lot we can do, right? You, you mentioned systems because I think this is important. What's important to me, and, and I realize this, I don't want this to sound callous for like the specific, you know, a specific hypothetical abusive case, right? That might have not been, you know, the, the, you know, mistakes can, can be made, of course. But I want a system like the system we have, which is that if...

• 43:18Speaker 02

  The government has some sort of reasonable cause and can convince people in a defined process to give data, yeah, the data should probably be given. What I really worry about is that we often live in a world where you can, basically, the government can instead ask, "Give me all your data, and I'm gonna look for problems. I'm gonna look for a crime." You know, the word "wiretap" comes from a time where they had to literally go to your house and tap the wire. And I don't think it was really thought of at the time, but this, the actual, physical act of having to do this had a barrier to the fact that you couldn't surveil everybody at once, right? It was just, it was simply, logistically

• 44:03Speaker 02

  impossible. We live in a world today where you can surveil everybody. So, we want to build services and systems where this is impossible, right? Where you have to, where the default should be privacy, as I said, privacy by default. And yes, we are also responsive to legitimate law enforcement requests, however you define legitimate. And that legitimate question will always be a matter of process, which we, of course, are only one player in this process. But I think that's the-

• 44:35Speaker 01

  There's legitimacy and there's like a market dynamic here. So, the Swiss government has applied a lot of pressure to Proton in the last few years. They wanted you to basically have an unencrypted VPN and be able to decrypt user data that was traveling over your VPN. I believe Proton threatened to leave Switzerland over this. And then you, you announced that you were gonna build a more distributed infrastructure and place some of that infrastructure in Germany and Norway. We're gonna come to chat control. At some point, the EU is gonna, yesterday, I think they passed a version of this law, and Proton's response was, "We'll just leave the EU. Like, we're gonna take ourselves out of this legal jurisdiction." I wanna talk about that stuff in detail, but just in the sort of broader context that you're talking about right now, you have to, you're picking, right? You're picking a

• 45:17Speaker 01

  foundation, and often the response is, "Well, if you change the legal foundation here, we will leave."

• 45:23Speaker 02

  Yeah.

• 45:24Speaker 01

  How real is that?

• 45:25Speaker 02

  It's dead serious. I mean, it's, with all due respect to Swiss authorities and everybody else, I think it would be absolutely suicidal to continue down this path. Part of the Swiss brand is privacy. It's been that way for 80 years. And, you know, they have a, they have a good via Proton largely, but also they have a good case for bringing that in, that reputation and that economic advantage, just because there's a lot of businesses, a lot of, a lot of commerce that requires confidentiality to the 21st century, to the digital age. And, you know, throwing it away is, is, I think, short-sighted, to

• 46:08Speaker 02

  say, to say the least. But it's a serious threat. The thing about digital services is they can be moved. There's a lot of flexibility in this. And if we get truly dystopian, there may be a world where there's no, where there's no place to move to. But at the moment, there are options. We hope to not have to do it. We hope that the powers that be are responsive to this and, and change course.

• 46:33Speaker 01

  That's a cost, right? I'm curious about this.

• 46:36Speaker 02

  It's a, it's a cost. Well, it is a cost. It's maybe not as much of a cost as, as you might think, right? We already have, we already have employees in different countries. We have, we have satellite offices. We have other things. We, we have data centers in Germany. We have, we have our data centers in Norway. You know, not to, not to be, not to be too blasé about it, but we could do a corporate inversion to somewhere else and then say, "Okay, all the legal requests have to come through here," right?

• 47:08Speaker 01

  How does that play with you can't sell it, right? The part where you're like, "It's a Swiss foundation. It's very hard to sell. You wanna leave Switzerland and reincorporate in Germany." Is the Swiss government gonna stop you?

• 47:21Speaker 02

  I'm gonna have to defer on that, because I'm not a lawyer. I have no idea. But I'm sure it's possible.

• 47:26Speaker 01

  Yeah. I'm just curious, because it feels like, at least as of this conversation, NATO still exists, Germany and Norway are still in the EU. Like, there's Swiss law, there's German law, then there's EU law. And the EU law is also getting increasingly intense about what they can scan, what can't they scan. I mentioned chat control earlier. That's the law that in the EU would require service providers to scan the contents of messages for CSAM material, for other kinds of infringing material. That's a big problem, right? I mean, a bunch of, I think Proton has said, "We will just leave the EU if you make us do this," right? This is gonna break the core promise. If you move to Germany

• 48:06Speaker 01

  and Norway, and then the EU passes the harshest version of chat control, are you geared up to leave?

• 48:13Speaker 02

  Yeah.

• 48:13Speaker 01

  Like, how ready are you? Can you pull the switch tomorrow? Where would you go?

• 48:17Speaker 02

  I can't. I, I, I don't know. I have to consult with other people for that. I'm not, I'm not deep in that. But, but it's, it's a real threat. I know it's a real threat. And I know we've made, we've made some preparations about where we could possibly land for this if, if both the EU and Switzerland become inhospitable, inhospitable to this. I, this is a, you know, there, there are separate things. There are the, there's the, there's the practical challenges that we, we all know this stuff is happening. There's a wave of whether it be age verification, or, or breaking encryption, or stuff like this, that this seems to be in vogue right now. And it's something we are fighting on the policy

• 48:55Speaker 02

  front. It's, in my opinion, very misguided. And we are trying, and hopefully, that at some point, the fever breaks and, and this, you know, you cannot brand a backdoor with an American flag and say that only good people can use it. And it doesn't work like that, or an EU flag or anything like that. It's, and, and maybe this comes a little bit back. So, there's the practical part. What do we do if this happens? What do we do if this happens? And, you know, at the end of the day, governments hold a lot of power on policy. And we, you know, you have to, you have to figure out how to, how you can comply. Or if you can't comply, you leave, you do something else, right? But there are a lot of countries in the world. And I think, I think we can, we can do, we can,

• 49:40Speaker 02

  you know, ultimately, we, we will do what we have to. The other part, just to, I guess, use, use your platform a bit to, to make, to make the case. This chat control, age verification, all this stuff, it's a very bad idea. And, and I don't wanna say, like, there are real threats to children online. And it's not that we shouldn't take them seriously. But there are ways to do this in, we talked about systems thinking before and systems design. There are ways to do this in that balance the appropriate concerns, that say, you know, can gate mature material behind age gates that don't reveal who you are, right? And

• 50:23Speaker 02

  once you build a system that essentially abolishes anonymity online, how long before that system, I mean, it's Chekhov's gun. How long before somebody comes along to use, if it's built, somebody's gonna use it eventually for purposes that it wasn't designed for? How long until China says, "Hey, identify all the dissidents for me," right? Who are using this? Because they have to use their, their, their ID for everything on the internet, right? We want to build systems, internet systems, that can't be commandeered like this. This is how we approach it. But I think this, this principle applies to the larger internet. This is a kind of a thought experiment, right? But there were some, I'm probably gonna butcher this, that this, but there was, there was some crazy statistic that a

• 51:07Speaker 02

  huge fraction of East Germans were actually employed as informants by the Stasi on the other East Germans, right? And, you know, in the height of the Cold War. The Iron Curtain, of course, fell pretty much right before the dawn of the digital age, in some ways, the internet age, right? I think you could argue, and I think you can look at counterfactual, or, you know, counterexamples like, like China. I think you could argue that had some of this, the Eastern Bloc authoritarian regimes made it into the digital age, that maybe they would have had enough control over information to not fall anymore, right? Because it's so much easier to do this. So, the fact that Facebook and Google, arguably, with their ad ecosystems, have built the most sophisticated, okay, China, perhaps,

• 51:52Speaker 02

  but the most sophisticated surveillance systems ever built, right? And we do the most American thing ever with it, which is we use them to sell, sell you crap you don't need, right? But that doesn't mean that's the only use for those. And the fact that those are sitting on the mantelpiece, like, like Chekhov's gun, waiting for somebody, waiting for somebody to pick them up and do something truly horrific with them, is a threat to free society. And all this other, you know, all this, all this other stuff with chat control and age verification is the same thing. It's saying, "We have this harm. Let's build a system to prevent this harm," that then can be used for, for really nefarious purposes. We need, we need to make

• 52:32Speaker 02

  sure that we don't engineer, engineer systems that threaten the existence of free society. Sorry, that was way soapbox, my soapbox speech. But it's, it's something I care deeply about.

• 52:44Speaker 01

  We need to take another quick break. We'll be right back.

• 53:19Speaker 01

  We're back with Proton CTO Bart Butler. You just heard Bart give a pretty impassioned defense of online privacy and why he thinks it's so dangerous to build systems capable of mass surveillance under the assumption that they won't ever be used nefariously. Now, I really want to dive into the tension that exists between that philosophy, which I broadly agree with, and one of the hardest problems playing out in politics and tech, where we draw the lines when it comes to child safety and the internet and what technology should or even can do to reduce harm.

• 53:47Speaker 01

  The reason I asked you so much about the structure of the company and its culture and how the systems are built are because that idealism is often expressed in Silicon Valley. I've heard it from all of the big companies that you have described. I hear it from big companies today. And then the compromises creep in. And sometimes the compromises are, well, we're domicile of the United States. We're just going to have to listen. There's nothing we can do. You can look at our warrant canary page to see how many legal requests we're getting, and that's going to be that answer. Sometimes the compromises are, look, we have to grow. We have to get bigger and that's it. And sometimes the compromises are, there's no way to build the system that would protect people the way we want and comply with the legal regimes. And I think encryption sits at the absolute heart of that tension.

• 54:31Speaker 01

  I'll give the example of Apple, because I think probably everyone is familiar with Apple. Apple routinely resists these calls for a backdoor. Right? And they're big enough to do it in the ways that they can do it, and then the iPhone gets zeroed out anyway. It kind of doesn't matter. And like, that cycle repeats in a way that it repeats. But, you know, if you talk to the folks at Apple, they're like, look, the regulators come to us and they're like, just be smart. Just do some smart stuff, smart guys, and find a way to do a backdoor that will preserve privacy. And Apple's response is, you cannot. We cannot nerd hard enough to solve this problem for you. I think there's some willful ignorance on the part of the regulators. I think the regulators know this. And then I think there's a mass of activists who want to protect children, who, maybe

• 55:16Speaker 01

  they do understand it, maybe they don't understand it, but what they certainly understand at a visceral level is the kids are being harmed. Right? And all of the other systems that everyone claims can ameliorate the problems or mitigate the risk to encryption do not actually exist, such that the kids are not being harmed at the rate they're being harmed today. You sit in the middle of this, right? The governments of the world come to you, they've asked you for backdoors, they've asked you for client-side scanning of chat messages to detect CSAM.

• 55:45Speaker 01

  What's your response to just be smart? Just nerd harder and figure it out?

• 55:48Speaker 02

  It's impossible to create a backdoor that can only be used by the good guys. And the consequences of backdoors being used by the bad guys are basically catastrophic, right? You mentioned, you know, there are still harms being perpetrated, you know, at these rates and that...

• 56:05Speaker 01

  Like at massive scale. I do think it's important to say that clearly. The harm is being perpetrated at massive scale.

• 56:10Speaker 02

  The concerns are legitimate, right? But we tolerate a lot of harms in the... We tolerate, you know, ingesting things that aren't good for you if you're an adult, right? And there've been, this is maybe US-centric, but there've been, there've been several, I mean, I don't know if it's still precedent with the way things are going, but there've been several court precedents that have basically said that, said that, and I think this actually has to do with scanning or otherwise, that have said that, okay, you, these must be compatible and balanced against restricting the freedom of adults to essentially be, you know, you

• 56:50Speaker 02

  can make society very, very, very secure by taking away all freedom whatsoever. Right? You can do that. This is a trade-off of what we want to make, but I don't think, but I don't think people fully internalize the fact that, that too much security is maybe a world that they don't want to live in as well. Right? Because it really, I think it was Ben Franklin, again, not to be, not to be too, too, too American, but I will, you know, people who would trade in freedom for security deserve neither. I'm paraphrasing. I am American, I live in Europe now, but I am, I am American originally, but there are a lot of people who really beat the drum of freedom in the, in the US, but are willing to essentially give away all, all privacy.

• 57:35Speaker 02

  And I think, you know, privacy and freedom are inextricably connected. You cannot be free without privacy. Right? So I think there's a trade-off to be made here. And we can discuss what the trade-off is, but the trade-off should not be that we make the entire system transparent to whoever wants to look. Because there are some really bad people who want to look. And even if you think that the government will never be that bad person, which, okay, seems naive, but you know, even if you think that's the case, there's all kinds of cyber criminals and everywhere else. And those backdoors that, or vulnerabilities, I guess vulnerabilities are more easily expected, but those vulnerabilities that have been found or introduced in cryptography have, have a long history of being exploited by bad actors, right?

• 58:20Speaker 02

  I mean, the, the history of this is clear. It's impossible to have a backdoor that can't be exploited by just, except for the people who it's intended for. So I, I think there are other ways to do this. I think we...

• 58:32Speaker 01

  Can you describe those other ways? I think maybe this is missing from the conversation, right? If there's other technical solutions to mitigate the harm, what would they look like?

• 58:39Speaker 02

  A lot of it would be getting in the weeds about how to...

• 58:41Speaker 01

  Go ahead. That's what the show is for.

• 58:43Speaker 02

  About, sure, about how to do things. But for instance, there's such a thing as zero knowledge proofs, right? Which, which, which basically involve the ability to prove that I am over 18 without actually giving you any other information about it. There's actually multiple ways to do that, some of which don't involve zero knowledge proofs, but these are a far cry from, say, you know, Discord got in trouble, I forget, a few, a month or two ago, from having a security breach, where, you know, they required, they did age verification, and then they had a, I forget what it was, some sort of open S3 bucket or something with tons of people's IDs and stuff? Like, this is not the right way to do it. But I guarantee you that people will do, that people will do this wrong. If every website has to collect your ID, we, you know, we, we are in deep, right?

• 59:28Speaker 02

  So there are ways to have issuers, you know, have, say, a trusted person issue you an ID. It can be a local credential, right? It can be a credential, and it can also be a credential that's stored in the cloud, but in, in an encrypted way, so the server can't see what the credential is, and then have your device selectively disclose age over 18. You know, the site that you go to requires that. It doesn't have to be, this is your address, this is your name, this is whatever. It can just be, oh yes, this person has a cryptographically signed affidavit that says age over 18, let them in.

• 1:00:03Speaker 01

  That's a technical solution to a regulatory demand, right? Do you think that the technical solution has to be written into the regulation? Or there has to be some regime of, this is how to do it the right way? Or do you think the industry has to come together and say, we're going to do zero knowledge proofs?

• 1:00:17Speaker 02

  I think it certainly helps that the industry comes, comes together. Like, there's a EU ID initiative, which is actually fairly good. It's not zero knowledge, but it's fairly good. There are some problems with this kind of credential. Obviously, digital credentials can be copied, so you want to make sure that it doesn't get copied and posted on the internet and used by a million people, right? So, so there's some anti-abuse trade-offs, right? You don't want one guy's ID from, from some country being used by every teenager in, in the US. But, um...

• 1:00:46Speaker 01

  I would have been that teenager, just to be 100% clear.

• 1:00:48Speaker 02

  Yeah.

• 1:00:49Speaker 01

  I would have absolutely been that teenager.

• 1:00:50Speaker 02

  Yeah, for sure. And I think in general, the history of writing, the timescales are just so different. The history of writing, um, exact solutions into, into regulatory frameworks is, is pretty dismal, right? But I think, but I think that the law could, laws that, that do this could write things in ways that mandate privacy protections in a way that I think has not been done.

• 1:01:17Speaker 01

  That would lead to the technical solution. This is kind of what I'm asking, right?

• 1:01:20Speaker 02

  Yeah.

• 1:01:21Speaker 01

  The government says, look, you have to start doing age verification. The kids are looking at all kinds of weird stuff online. We have to start gating some of this content the way that we gate, I don't know, porn stores physically. Okay. Where we gate R-rated movies. Okay, we got to do it. And the industry is going to sort of inevitably pick the cheapest solution. We're just going to store driver's license in S3 basket.

• 1:01:38Speaker 02

  Sure. Or they're inevitably going to, going to pick the solution that allows them to market that information in some way.

• 1:01:47Speaker 01

  I guess this is like, you're sit at the middle of it. Your ideals and your structure prevent you from doing the cheapest, worst version of this. How do you make the industry match your values? Is it the regulator mandates a solution, which you seem to think is a bad idea? Is it, they put constraints on the solution which lead everybody to do the right thing? Where does that come from?

• 1:02:06Speaker 02

  If I knew the answer, I would, I would, I would be pushing it. But partnerships are one of them. I think, I think certainly regulatory guidelines about what you can store, what you can't store, how much you can know about your customers is probably, is probably key. This gets into maybe competition stuff a little bit, but a lot of our, a lot of potential harms also comes from full vertical integration of a lot of stuff. So, you know, mandating that things be separate entities and thus, I'm not going to flush that out, right? And this type, but, but, but this can do, this can do a lot to prevent certain types of harms and temptations to abuse stuff.

• 1:02:46Speaker 02

  But I think regulation and having privacy requirements in the regulations is certainly the, certainly the first step. And that, you know, compliance, compliant implementations there at least have that baked in. I think also, and this is hard, pushing it on the operating system manufacturers is also, which is, it can also be kind of dangerous because it helps entrench those, those choke points, right? We already have a lot of choke points where Google and Apple, you know, have these...

• 1:03:19Speaker 01

  It's funny, Apple really is pushing against this, but Apple loves to be entrenched as a choke point. Why do you think they're pushing against being the age verifier?

• 1:03:27Speaker 02

  It's a good question. I don't actually know. You'd think that they would jump at it, right? I mean, they, they certainly love having, being the choke point of the App Store and charging everybody 30%.

• 1:03:36Speaker 01

  I know you have been in a fight, like Proton has been in a fight with Apple over App Store policies. This seems like one where their interest would be obvious, to say, actually, we will, we will maintain control. This, this 30%, like, stop bothering us about the 30% regulators, because we will provide you age verification.

• 1:03:54Speaker 02

  They might view it as simply a no-win game, in the sense that you can KYC, know your customer as hard as you want, but there's going to be something that slips through and then you're responsible. So maybe they just want a third party to do it, so that's not their problem, right? I don't know.

• 1:04:11Speaker 01

  That might be the whole answer. Let me ask you, we talked about age verification. I'm not sure there's like, an answer, right? That will solve every problem. But you've laid out some technical approaches that will at least balance the harms. When I said their harm is happening at massive scale, what I meant was CSAM, right? What I meant is child sexual abuse material. That is happening at massive scale over the internet. We all know it. There's lots and lots of ways to mitigate it. And then there's just platforms we can't see into where it's going to happen anyway. Proton is one of them. iMessage is actually another, if you fully encrypt iMessage. Apple, it's a lot of criticism for that. You know, their, their response is the same as Proton's like, there's no way to do this. We, we simply cannot give you a way to do this. It does not create the backdoors for all the other bad actors you want. We're not going to do it. What are the solutions to mitigate the harms of

• 1:04:56Speaker 01

  CSM without creating a backdoors? Cause I feel like that is also missing from this conversation, especially when I talk to the activists who are laser focused on the scale of the harm.

• 1:05:06Speaker 02

  I can't describe exactly what Proton does because, you know, anti-abuse is one of the, one of the things where security through obscurity really does actually help, right? It helps that the bad actors don't know what we do. But I can say that you can fight CSM without content scanning, like it is possible. And, um, a lot of that is, uh, anyway, I probably shouldn't say, but you can, you can fight CSM. I'm not saying it's the same as, it's not as effective as scanning everything, but also in the age of AI image generation, who knows what's real or not. Not that AI generated CSM is good, but the point is like, you don't, you don't know if there's a real victim, is there

• 1:05:50Speaker 02

  not, et cetera. So I think there are alternative things to scanning every image that you can do to fight CSM. And we've done this for years and it's hard to know how effective we've been at it because we don't really know what the denominator is, but we have, we have been, I think, uh, at least in terms of networks that we've identified, like it's, it's, we've identified some and we've shut them down and we've, you know, we've, we've mitigated this. We have a strong interest in keeping,

• 1:06:20Speaker 02

  in keeping bad actors of all kinds off the platform. The mission's not going to be served if it's, if it's, oh, this is a platform for criminals. And as a result, it's not. We've put nearly 10% of total company resources to fight abuse. We are dead serious about driving, you know, making abuse as little as possible on the platform. So it can be done. It's a cost center and one that we eat, right? It doesn't make us any money, but it can be done. I think the other thing is, you know, the digital sphere isn't the only place where these crimes are committed, right? And, you know, there's the physical sphere too. There's the actual victims. There's the actual people harmed by this. And I think, uh, you know, everybody wants their job to be easier. Cops are not accepted from this, right? So they would love to scan everything on

• 1:07:04Speaker 02

  the internet and I don't actually blame them for this, for this request because it would make their jobs much easier and they, you know, they really do want to reduce harm. However, I think we need to make a trade off between that and the harm, you know, the threat that is to free society as well. So I think that there's probably more that can be done in the, uh, in the sort of physical space as well to fight these kinds of, these, these kinds of abuse with resourcing and whatnot. But I will say, yeah, I'll, I'll repeat, content scanning is not the only way to do this. And there's also some content scanning that can be done, uh, without violating. And, you know, it can be client side. It can be other things that can be done in, in, in secret, in, in ways.

• 1:07:47Speaker 02

  This is very fraught too, but there, there are things that can be done to do this that don't violate privacy at a massive scale, which is a price that I, I think is simply too high.

• 1:07:57Speaker 01

  You're saying there's a system you can't quite describe that is effective at, at stopping the harms of CSM at scale. Is that verifiable from some of the people that wish to impose the regulations? Is it auditable? Security through obscurity has these problems, right?

• 1:08:11Speaker 02

  Yeah.

• 1:08:11Speaker 01

  We have to, we have to trust you. A lot of this conversation has come back to how much we can trust Proton structurally, personally.

• 1:08:18Speaker 02

  The problem is I don't know what the denominator is. I don't know if we found 50% of them. I don't know if we found 10% of them. I don't know if we find 1% of them because all I know is the stuff we found. And even then it's a probabilistic, you know, it's, we don't see the images, right? Sometimes we have a better clip, but I won't say how. But, um, but, but, I mean, but we don't see the images in any case. We are not legally capable of doing that. And God knows, we don't want to subject our employees to that anyway. But no, we don't know what the denominator is. So I don't know how effective it is. I, um, I do know that, uh, that, uh,

• 1:08:54Speaker 02

  that we can correlate this to some degree with the kind of legal requests we get. And that those are pretty few and far between, right? I would think that if this were, you know, if this were a massive problem on the platform, that we would get a lot more legal requests for metadata regarding this. That is our proxy, not for CSM in particular, because we often don't know where the requests come, but in general, one of our proxies is, okay, how many legal requests are we getting for data as opposed to, you know, how much, um, how good we are at anti-abuse, right? If legal requests go through the, the roof, which they haven't, then obviously there, we have a big problem with anti-abuse. And that has not been the case. So, um, I know

• 1:09:38Speaker 02

  that's a kind of a wishy washy answer, but the data is private. We can't read it. That's, that's our entire, that's our entire thing. We also have a reporting system, of course. You know, people make mistakes like any other. So if, you know, if you, if they share an image, which is, this can be reported, that, that, that can go. So there are lots of mechanisms. Again, I'm sort of tiptoeing around saying any details because I don't want, I don't want this stuff to not become effective, but, uh, but we do a good job, I think, um, relative to external indicators that we can see. And, um, and yeah, we, uh, I'll give you an example where this is, it's not CSM, but it's a similar thing. Okay. We, we used to get a lot of requests.

• 1:10:24Speaker 02

  Relatively speaking, a lot of requests for ransomware accounts. Okay. Those people go to Proton, they said, the, the, the, a Proton email that you would, I don't know, with a Bitcoin something, right? Whatever. We got, and we would get this in the legal requests, right? We'd get this and we'd, we'd be able to look at that and say, oh, that's definitely a ransomware account for various reasons, right? We got really, really good at killing ransomware accounts. And now when we still occasionally will get, we'll get them, but they'll have been disabled by us for abuse six months before we get the legal request, right? So this, I'm using the ransomware as an analogy, but we do have kind of a feedback loop.

• 1:11:08Speaker 01

  So there's something in your system. There's some set of indicators that you can detect and you're just not going to tell me what they are. But there's some set of indicators of how an account operates in your system that lets you know what it's being used for?

• 1:11:20Speaker 02

  Yes.

• 1:11:22Speaker 01

  Has any government ever asked you what that set of indicators is?

• 1:11:25Speaker 02

  No.

• 1:11:26Speaker 01

  Interesting. I'm just, hopefully that doesn't happen today. Cops, if you're listening, forget, forget that you heard any of this. Uh, we're running out of time here. I want, I do want to quickly at the end, ask about AI because we've talked about systems and dynamics that I think are pretty familiar. Like I came up on Usenet and Slashdot, and I probably heard that quote about liberty and security 10,000 times. Like that is the foundation of my life on the internet, right? Is like this debate from, I don't know, the eighties and nineties up until now is the same debate. And maybe AI is going to like flip over the whole apple cart, right? We can now do cybersecurity at scale in ways that governments are stepping in and stopping the models from shipping, whether or not that is correct or not. That is the thing that is actually happening in the world. We can generate vast amounts of synthetic data that might trip all your detection

• 1:12:09Speaker 01

  systems, whether or not any harms are actually downstream of that data. You've shipped an AI assistant because a bunch of countries do not want to rely on American model companies and having data sovereignty in Europe is important to them. That seems like a market opportunity for you. What is your approach to all of this? Are you reliant on the frontier companies? Are you building your own model? How do you protect your data from them? All of this, at the very end feels like it upsets in significant ways, the structures and the systems we've been describing up until now.

• 1:12:40Speaker 02

  I don't think it's had as much of a dramatic effect

• 1:12:46Speaker 02

  on the topics that we've discussed. We have Limo, our own internal, um, which is run on, we control the systems that it's running on. And, and this is, this is our, we want to be independent of third party models and things, and, and things like that.

• 1:13:01Speaker 01

  Is that actually a model you've trained or is that reliant?

• 1:13:03Speaker 02

  No, no, it's, it's, it's open source models that we've, that we've compiled. And it's a collection of open source models that we sort of stitched together based on their strengths and weaknesses. But yeah, we, I mean, we, we don't have a billion dollars to set on fire to train, to train our own models. So, and, and this is actually why, uh, we might be one of the few. The few AI companies that actually makes money doing it, because we don't do any training, right? We just sell the inference. I mean, not that there's not a lot of work involved in doing, in doing, in building Limo. But, but yeah, we don't train our own models. I guess there are two, two different sides. There's, there's the, how is AI going to affect the industry in general? And maybe how it's affected, how it's affected us and

• 1:13:47Speaker 02

  how we use it. So for a lot of our stuff, in particular, the client side stuff is up in sorts, right? So we've, we, and the client side stuff is not user secrets either. So we've also, we've tried out, we've tried models from, you know, from Anthropic and ChatGPT. We also, we've also tried Limo and OpenAI, because a lot of it's public anyway, and the, and it's not user secrets, right? So we're going to use the best tool to make it, to make us go as quickly as possible. And it has changed software engineering, probably permanently, right? A lot of the, in terms of the, the amount of time spent on, you know, actually writing code versus the other things. But, but maybe not so fundamentally.

• 1:14:29Speaker 02

  I mean, people talk about this SaaS apocalypse thing, that every company is going to be writing their own enterprise software, and I just don't buy it. There's a very big difference between being able to cook up, you know, your own custom, purchase order software, ERP, or whatever, and that works for your small business, maybe sort of, and the, the kind of things, whether it's compliance, whether it's, whether it's scaling, whether it's reliability, all these things which enterprises need. I just can't, cannot see every company building, you know, oh, software is dead. We're not going to have vendors to do this that can promise us things that we need, right? The other thing about this

• 1:15:11Speaker 02

  is, to me, if anything, it's, it shifted what is valued in software engineering towards more senior level skills, perhaps, right? But it's not fundamentally changed what good software is, or how to architect good software, whatever. If anything, things like reviewing other people's code. Well, today you're reviewing other people's code, and you're reviewing the robot's code. But it's kind of the same skill, right? I mean, so what maybe the balance before was a little more towards, okay, can you write code very quickly, accurately, with, you know, this kind of thing? Where now it's like, okay, it's more towards, can you review code very quickly? But that's, which is maybe a more senior level skill.

• 1:15:55Speaker 02

  But it's still part of the software engineer toolkit, designing good software, what good software means, what good software looks like. One thing that we found, I think, is that LMs work a lot better when your code is well architected. Like, if your code is well designed, the LM will actually work better. If your code is a spaghetti mess, the LM will not work as well, because they're, you know, they're logic engines, right? There's more chances for them to get confused. So a lot of the fundamentals of building software, yes, the industry has changed, but a lot of the fundamentals of building software are different. The other thing we found is that, you know, your, the time you spend writing code has gone down.

• 1:16:35Speaker 01

  Yeah.

• 1:16:36Speaker 02

  Now there are other bottlenecks in your pipeline that you, that you have to do. Now maybe your CI takes too long. Now maybe your other things take too long, right? So it's, it changes maybe what kinds of things you need to optimize. But at the end of the day, for, I think you, for most senior software engineers, the amount of time that they actually spent writing code was, was frankly, not the, not the, either the hardest part of their job, or what they spent most of their time on, right? Software is probably the most, the most affected industry, and the most tractable industry for, for LMs that I've seen, at least so far. Obviously, there are other uses for it, but you know, it's

• 1:17:20Speaker 02

  software, it's, it's rule based, you can test for correctness, you can mitigate a lot of the things that LMs might kind of go off the rails and hallucination. At the end of the day, you need working code, right? So you can test this kind of stuff. So you can mitigate a lot of the, a lot of the problems with LMs. Software is your best case scenario. And even then, from what I've seen, it's, it's an evolution. It's a, it's, it's not a, it's not a revolution. It's definitely a big change, but it's, it's largely, it's a productivity change, not a, you know, not a dump everything, start over. Maybe I'm wrong about that. Maybe as models improve, it'll get more and more dramatic. But, but at the moment, it's, it's a hell of an opportunity for productivity. But I think the fears,

• 1:18:05Speaker 02

  and software is what I know best, but I think, I think the fears are a little overblown.

• 1:18:10Speaker 01

  Let me ask you on the initial of that, right? The, you mentioned earlier, now it is possible to look at all the emails and do mass surveillance. Yeah, Anthropic rolled up to the United States government. It's like, you can do this with our model. We don't want you to, because we don't, we don't actually think it's good enough. And we think there's like, real problems with doing that. You're faced with that as well, right? You have your own internal AI model, there, maybe your user are probably generating more data inside of your systems than ever before with AI. It just seems like a thing that happens when you add AI to a system. The models get better when you feed them more data. They have a voracious appetite for data. That's a lot of pressure on, we're going to keep everything private, right? It's a lot of pressure on who gets to see your stuff, and why do they get to see it? And what can they do once they have it? And the value of having it seems to be going up, right? The value of collecting all the data seems to go up.

• 1:18:54Speaker 01

  How do you respond to that, even as you have to roll out AI systems in order to compete with the big tech that is putting it literally everywhere?

• 1:19:02Speaker 02

  I mentioned before, is, you know, privacy is control, privacy is sort of self determination in the sense that you control your data, and you control who it's shared with. And maybe that who is an AI system, right? And maybe, and that's okay. But, and I think as long as this is sort of an opt-in thing from the user, there, you know, this is actually an internal debate, but there, there are, we already have features, like, for instance, okay, we have a mailing list feature, okay? I promise this has something to do with.

• 1:19:35Speaker 01

  Yeah, it's a big build. Here we go.

• 1:19:38Speaker 02

  We have a mailing list feature, and that feature is done in a way that it is, it is private. I send you an email, or send the mailing list an email, it sends the email to everybody else. The system doesn't see it, okay? But if you want people outside Proton in your mailing list, then we can't do end-to-end encryption anymore. We have to turn it off, right? So when you add external people to your mailing list, you're prompted, hey, do you want to turn, you know, external recipients, you need to turn off end-to-end encryption? The user says yes? Okay, because that's part of their workflow, they need it. And at that point, we still don't save a copy, right? Everything saved on Proton is encrypted at rest. However, we do have the clear text email going through our system

• 1:20:21Speaker 02

  because we need to send it out to whoever the external recipients are. That's the kind of model I see in the future with some of these. And it's not just AI systems, it's also integrations, and especially as we get more into business clients and stuff. I need this integration into my CRM, right? I'm going to choose to share this mailbox with my CRM. That's okay. You know, that, that can be a decision. And they, and we, it's, it's our job to build a user interface which communicates what the consequences of that are, right? But, but privacy is fundamentally, it's not about not sharing stuff, it's not about not sharing your photos, not about not sharing your data with, with third parties.

• 1:21:02Speaker 02

  It's about sharing, sharing the data with third parties, who you choose, right? And not just by default with everybody, which is sort of the, the paradigm that's been the big tech paradigm for, for Web 2.0, or whatever you want to call it, the last, the last two decades. So I, I think that, I think we can reconcile this with the Proton thing. Because it was never, it was never just about the encryption. Like, the encryption is, is a tool to the end. It's all about, hey, the fact that I put, the fact that I stored data on somebody else's computer on the internet does not mean that I give them control to do whatever, whatever they want with it, or that I trust them not to lose it. So we're

• 1:21:45Speaker 02

  going to try and build a system where I retain that control, and I can still participate, you know, in modern, modern services, modern online services, right?

• 1:21:53Speaker 01

  I feel like we're going to have to have you back soon to see how that's all put to the test. Like, this is new.

• 1:21:58Speaker 02

  Yeah, yeah.

• 1:21:59Speaker 01

  And I feel like we're, in the next couple of years, we'll, we'll find out more. So we'll have you back soon. You've given us so much extra time. Bart, thank you so much for being on Decoder.

• 1:22:06Speaker 02

  Thank you. Thank you so much. It was a blast.

• 1:22:10Speaker 01

  I'd like to thank Bart Butler for taking the time to speak with me, and thank you for listening. I hope you enjoyed it. If you'd like to let us know what you thought about this episode, or really anything else at all, drop us a line. You can email us at [email protected]. We really do read all the emails. Or you can hit me up directly on Threads or Blue Sky. We're also on YouTube. You can watch full episodes at DecoderPod. We also have a TikTok and Instagram. They're also at DecoderPod, and they're a lot of fun. If you'd like to cut or please share with your friends and subscribe wherever you get your podcasts. Decoder is production The Verge, part of the Vox Media Podcast Network. The show is produced by Kate Cox and Nick Statt. It's edited by Ursa Wright. Our editorial director is Kevin McShane. The Decoder music is by Breakmaster Cylinder. We'll see you next time.

• 1:22:45Speaker 03

  New and exclusive holiday decor just launched online at The Home Depot. Check out the viral Grand Duchess collection with smart holiday lights you can control from your phone. Easily customize the twinkle across trees, wreaths, and garlands to match your holiday vibe. Create stunning light effects, then switch it up with the tap of your fingers. Even the man in red will be impressed. Shop the holiday decor you want online only at homedepot.com.