Is AI A Privacy Disaster? And How To Fight Back. — With Andy Yen
Channel: Alex Kantrowitz
Published at: 2026-03-25
YouTube video id: 63-pbAxcLfY
Source: https://www.youtube.com/watch?v=63-pbAxcLfY
Could AI expose us to a whole host of new privacy concerns? And what should we do about it? Let's talk about it with Proton CEO Andy Yen in a conversation brought to you by Proton. Andy, great to see you again. Welcome back to the show. Hey Alex, it's great to be back. And yeah, I think a lot of digging to here AI and privacy. Definitely a hot topic these days. That's right. You know, we spoke a few years ago before the LLM craze took off and I really started to thinking of you recently when I realized what was going on when I'm using applications like ChatGPT and Claude and Gemini and what was going on in the background. So, let me just give you an example. Uh I recently found out that we have all been opted in to AI training when we use ChatGPT and Claude and only if you go in and toggle off those settings can they not use the conversation that you're having with those chatbots to train future models. That kind of astounded me, to be honest. I mean, I was definitely having some pri- what I thought were and I probably should know better at this point, but what I thought were private conversations with these chatbots. And the next thing I know, the material can be used for training. So, I'm curious to hear your perspective about that and what do you think it says about the state of digital privacy today that that can happen? Well, actually it's worse than you say. It's not just it can be used for uh training. Uh even if they don't use it for training, they can still store it uh for a long time, actually forever. Uh they can still hand it over to a government under a subpoena. Uh they could uh leak it if they're hacked in the future. Uh so, just because you've opted out of training doesn't mean you opt out of data collection. In fact, all the information is uh still there uh you know, on Google, on OpenAI, on you know, Anthropic's servers. Uh so, it's actually quite a bit worse than you imagine. Uh it's opting out is one thing, but it doesn't make mean that they don't collect it. So, this is I think the first point I would make. But, to kind of your point, uh yeah, I think LLMs to you you to it frankly, it's a privacy disaster. Um And what is LLMs actually for most people, it's kind of like a more efficient way for humans to communicate with computers. This is This is all AI actually is, right? People say AI is this huge new thing. No, it's just it's a continuation of a trend uh past 30 years of being more efficient. But efficiency comes at a cost. Uh you know, a lot of people today in fact use AI almost as a replacement for search engine. It's like a conversational search engine. The difference is when you use search, Google knows what you're interested in, but they don't really know that much about you. When you're having a conversation with an agent, uh yeah, with with AI chat, and you're speaking to it for months and months, and you have uh tons of data being collected, well, now it knows what you talk, the way that you converse. It has a much deeper insight into your personality. Uh and this is why I think the information that uh AI has, like if you share that information instead of Google search, it is really supercharging Google's ability to understand who you are, and it knows a lot more than if you were just using a search uh product. Uh so, it's You can think of it as Google's business model, but now on steroids. So, let me tell you one more thing that I do to try to defend myself, and you can tell me whether or not this is foolhardy. Mhm. There are There is an option uh both in OpenAI's ChatGPT and Anthropic's Claude where you can toggle to an incognito conversation. >> Mhm. Does it Does that protect me at all? Mhm. Well, uh Google Chrome had incognito as well, uh but then, if you've been reading the news, Google paid probably several billion in class action lawsuit settlements, uh and that basically showed that incognito wasn't as incognito as it was claimed. I At the end of the day, you're kind of taking their word uh on whether it's incognito or not. Uh and past history shows that uh, when Google says something is incognito or anonymous or you know, uh, uh, private, it generally hasn't been the case. Uh, and and this is why I I think uh, anything that is any like trust-based model of privacy uh, has its limitations. So, it's probably better than uh, you know, using the default setting as you say that allows training and all that. Uh, but I wouldn't consider it to be let's say bulletproof given Google's uh, track record here. So, I did upload my tax returns in this secret browser. That was a mistake. Yeah, >> secret chatbot. Yeah, you probably shouldn't have done that. Maybe you put it in Proton Lumo, that would have been better, right? Okay. Yes. So, Proton does you have your own LM called Lumo. We're going to talk about that uh, in a minute, but um, but basically your perspective here is um, when we're using these these bots, don't trust them. Don't trust them to protect our privacy and that is scary because we tell, you know, speed Yes, they're more efficient, but I think you've already pointed out this. We tell these LLMs more uh, than we tell any technology platform about ourselves. By by like orders of magnitude. Mhm. Yeah. Well, it's a bit like if you have a conversation with someone in real life, you will know them a lot more than you know, someone that just looked at what you ordered on Amazon for for instance. Right. And as humans, we actually get most of the insights about people through conversations. So, just imagine like you know, you and me talking, right? Let's say we meet at a pub somewhere. I know so much more about you cuz I met you in person and had a conversation with you versus if I was just like you know, reading your LinkedIn or or or something like that. Uh, and this is the element that we'll miss is that in fact, AIs are going to get better at discerning who we are than probably even a psychologist. And some people use these AI's as psychologists in fact. Right. I mean, there's an argument to be made that they will know us better than a friend would because we'll tell them things that we won't tell Yes. Exactly. Yes. And that's scary when you think about it. It's it's it's completely scary. Definitely. Now, another issue that's come up around the AI world, but I think it's actually something that's been a pervasive concern across all tech platforms is the use of these platforms by kids. And we know that kids have been using them, you know, LLM's in particular, but also social media. There are currently trials going on about social media addiction and what social media leads to. And of course, there's worries that as kids get more engaged with chatbots, who knows where they're going to end up. And so Proton, which again does encrypted email, still doing docs, calendar, LLM has done a survey of parents and asking them about the way that they feel about their kids use of technology. And I thought it would just be a good opportunity since we're talking for you to share a little bit about what you learned and what the parents told you guys. Yeah. Well, it's quite interesting because there are so many people who, you know, today just give kids phones and then, you know, kids have always also discovered AI. In fact, I would say the younger generation is maybe even better using AI than some of us, right? And so just give you some top line numbers. Uh Today around 70% of kids have access to a smartphone by the age of 10. Right? This is this is like seven in 10. Uh and of those actually around three quarters are using Gmail. And that's a staggering number if you if you think about it. And a 10-year-old child doesn't really understand privacy. It doesn't really understand what ChatGPT and OpenAI is doing with their data. They they have no clue whatsoever. It's not really taught in school, uh but they're already on these platforms. And I think parents are starting to realize sort of the risk that comes with this. So, if you look at uh the survey that we ran, if you look at uh if you ask parents what they think about all this, well, 41% of parents say that if they could start over and do things differently uh with putting their putting their uh children online, uh they actually would want to, you know, uh start over. Uh and 60% uh would want to be able to erase all the information uh that about about the kid that today is on these big tech platforms. Uh now, it's obviously a little bit too late now for many of these parents, uh but it shows that there's a huge amount of what what we call parental, you know, uh you know, uh regret. And it also leads a lot of anxiety, right? I I you know, if you look at parents, uh around 80% uh are really concerned about uh you know, their kids uh you know, online privacy. Uh and these are sort of uh massive numbers if if you really think about it. Uh and and this is I think sort of um quite surprising, cuz a lot of parents in the last couple of years, they've just sort of sleepwalked into the situation. They're now starting to realize, "Oh maybe we shouldn't have done that." Right. And so, but let me ask you because sometimes there'll be platforms like Facebook or or maybe Google as well, that will say, "You have to be like X years old to use our platform. Do you certify that you are?" Some are even saying, like I think OpenAI has been saying recently that we have technology that can determine, you know, smartly whether you're a minor or not, and then prevent you from I think the use case was using adult mode in ChatGPT. Uh >> [clears throat] >> do you trust those platforms with that sort of those guardrails or no? I think the the guardrails could work and it's good that there are some guardrails in place. But ultimately, the business model of these platforms is to extract the data of our children. This is what they are all about. And in some sense, the guardrails are running counter to their business interest. And we have seen historically that when there's this conflict of interest between what is good for the business and what is good for society, we know the choice that Zuckerberg is going to make. We know the choice that Google has made in these things. So, I wouldn't really trust the guardrails put in place by a company that doesn't really have the incentive to put in guardrails. But one of the There was this social media trial. I think it was a meta executive that said, "We make the least amount of money on our young users." So, that would maybe suggest that they don't wreck as much. Uh He He didn't say He didn't say we make no money, right? He said we make less money compared to parents. Right. Uh but it's about building habits. It's about getting children What is What is social media? Uh I I I view it literally not so much different than a neighborhood, you know, part or crack dealer. Uh it's there to get kids hooked and addicted and put a lifetime of of an dependency. Uh and yes, children may not be your most lucrative uh audience today, but uh they do get older, they do have more purchasing power, and in 5 10 years, they become sort of your uh you know, uh the cow that uh you're going to milk for profits. Um so, in the long game, yes, they do want to get children hooked because that is how they're going to turn profits in the future. This is a building out the customer base that's going to buy their crack 10 years from now. Yeah, I I I don't know if I'm fully on board with the crack example, but I will say this. Uh it's like it's obvious that when you're a kid, you're not going to have the purchasing power that an adult would. So, to come out and say it's some saintly thing that you don't make as much money on, let's say, under 18 as you would, you know, 25 to 40, well, obviously not. Yes, but it's also a long game. It's about getting an entire generation completely addicted so that they can monetize them in 5 10 years as they get older. Exactly. So, so how do you think, you know, we've talked about a couple of issues here. Talked about uh and privacy. Talked about kids use of of technology and the problems there. Any solutions that you might have to that? Well, fundamentally, it's a business model problem. And if your business model is ex- is exploitation of data and abuse of privacy to generate money, uh that conflict of interest is always going to exist. And what Proton has tried to do differently is have a business model where the only incentive that we have is to protect user privacy. And this was really the motivation behind, you know, launching Lumo last year, which is sort of the private AI. The way that we describe Lumo is it's the only AI where your conversation is actually private. Uh and this is ensured through, you know, very strong encryption. In fact, uh we don't have any ability to go in and query past conversations. This is different from what Open AI can do. This is different from what Google does. Uh and this is because our philosophy is the best way to protect data is not have it in the first place. And that is only possible because we have a business model that doesn't require us to get your data in order to make money. Uh so, it's really a business model problem. And that is also the solution. And this is why, no matter what Google and Open AI say about your privacy, they simply don't have a financial incentive to to protect it. And Proton is a very different business because that is the only incentive that we have. The only reason people pay us anything is cuz we we actually put out privacy. And without that, uh we wouldn't have a business. And this is different because it really aligns our interests with the interests of the customer. Uh and this is in my view, the foundation to a more ethical and let's say uh more responsible internet. All right, but maybe it's maybe cigarettes is the better analogy, right? It's just you know it's bad for you, you keep doing it. Yeah. You need an alternative, but it seems like there's no way out. Mhm. Yeah. And and and and for us, uh there is clearly a way out, right? You know, um the whole point of the Proton ecosystem is it's it's a way for you to opt out of big tech data collections, uh you know, a framework. Uh today, if you look at email as kind of example, uh email people think about as communication, but really what it is is identity. It is the essence of who you are. It's the thing that connects everything online, you know, around you. And the way to look at this is if you go back to the history of like Google, the first product was of course, uh you know, um search. But what was the like second product? For for the ones who are older and remember this, uh you'll remember that the next thing that Google released uh was not all the other stuff they have today, right? Actually it was email. Uh Gmail was probably Google's second product, uh you know, maybe maybe third if you count the ads uh product that they put out as well. And why did Google go so quickly uh into email? Uh the reason is because they figured out in order to get all your information and correlate it to single profile, they need you to be logged in. And the one thing that you're always logged into all the time online is your email account. So what email actually is, it's a tool that allows them to have a profile linked to your real identity uh that gets all the your you know everything you buy, everywhere you travel, all your communications. But not only that, it forces you to be permanently logged into Google's ecosystem. So when you go to any website that has Google ads, uh the Google cookie is there. When you go to any website that has Google analytics, and this is like 70% of the web, uh all that information gets added into the profile that Google has on you. Uh and the reason uh Proton we started with email is because the way that you opt out of Google's ecosystem is actually by logging out of Google. And to do that, you need to have something to replace Gmail. And And this is why uh you know, in our view, uh ProtonMail was the first product that was the most effective at letting people opt out of Google's ecosystem entirely. And so then going back to the kids conversation, if you I'm just going to like throw this out there. I was just speaking with a friend who's like, "We got our baby an email address." That is one of the more consequential decisions you're going to make for a child in terms of well, if they are they going to start leaving this digital footprint from the time they can type? Yeah. And so sounds like you have a counter to that. And I you know, that Proton is going to make email address is going to enable parents to reserve email addresses for their kids. And I think it's a pretty pretty compelling package that you're offering. >> Yeah, I'm happy to talk about this. Uh today at some points your child needs to have a a a one existence. And what is an email or even any account? Uh you know, the your email it's more or less your digital passport. And I would argue in the 21st century the more important passport than even your physical passport. Yeah, like I've already give you the option between losing your passport versus losing your email. Actually, losing your email is probably more painful than you know, losing that passport. Uh so what you're doing when you get, let's say, a Gmail account uh for your kid? Well, you've gotten them an email identity, but you've also created for them an advertiser ID, which is going to be linked to them for uh the rest of their life, and is going to opt them in to Google's master surveillance and data collection ecosystem. Uh, so, all the things are wrong with the internet today, all the things that, you know, have caused so many problems to to to society, to democracy, to our privacy, you're essentially enrolling your uh child into this super shitty ecosystem that in the future is probably going to come back and ruin their life in some way uh 20 years, 40 years down the line that you can't even foresee and predict. Uh, and what we want to do at Proton is give an alternative. Um, you know, every day, and this is our data, right? 76% of uh parents, the first email they and the first account they give their uh children is Gmail. And this is really a mistake. I think uh you cannot be a responsible parent today, know what Google does, and willingly sign up your child into that ecosystem. Uh and so, what we want to do with Proton is, you know, we have this uh new campaign around the idea of born private. A- A- And what is born private? It's simply the reflection that all of us when we're born actually were private. No one is born into Google's ecosystem. Uh it is you're added in through sort of, you know, bad choices made by your parents, more or less. Uh so, what if instead of having children default into Google's ecosystem, there's an alternative where you can actually get, you know, a Proton address uh for your child, reserve it today, and opt them out of, you know, uh Google for the future. Now, reserving email address is quite uh difficult. If you look at, uh you know, Reddit conversations, uh actually the suggestion is, "Hey, if you just had a newborn, please go create a Gmail account and then log in once a year so it stays active and give it to your kid when they get old enough, right? Now, um that's number one, kind of a shitty user experience cuz you have to log in every single year and and keep that account alive. Uh, so Proton's uh Born Private program is just it's completely different. It's first of all, you're not in Google, so you have avoided the entire Google ecosystem. Uh, you get a Proton username, but we but actually we we reserve it for you 15 years. So, you don't have to go, you know, um every 6 months, every 1 year log in continuously for 15 years in a row. Uh, you just get your Proton account for your child once. It is valid for 15 years. And at any point in the next 15 years, you can activate it uh or, you know, um your children can can activate it. Uh, and all and and and actually uh in exchange, uh you know, it's not going to be completely free because things are free can tend to be abused. Uh, so there is a uh $1 fee. It's a very, very small amount of money. Uh, but it's a $1 fee to reserve an account for your child for 15 years. And it's really giving them a future outside of the Google ecosystem. Right. And and you at Proton have not only email, but competing uh services like Docs and Calendar. Yeah. And um Yeah. So, it's it's the whole thing. Uh, it's uh you know, um email, uh calendar, file storage, uh Docs, uh you know, Sheets as well. Uh, also an AI that is uh you know, um competing with Gemini, but of course private. Password manager, uh you know, a VPN service. And I would I'll bet probably 15 years by the time, you know, your kid is ready to claim that account, probably a lot more in the ecosystem. Uh, and so, our goal in the long run is to provide an alternative ecosystem that is privacy-first, user-first, uh that is not built on an advertising and surveillance business model. Uh, and you know, uh I the born private campaign from Proton this born private program it's just a way to make sure that the next generation doesn't fall into the same trap that our generation fell into. Right. So you won't do ads you've never thought about doing ads? No it's just not our business model it's not it's not consistent with our business model. Actually you know ads is only effective if you can very effectively target and show the relevant things to you know your audience. But because everything in Proton is encrypted we don't have the ability to go in and read your emails. We can't tell what you're interested in what you've bought if you're male or female if you're a gay straight you know if you're a Democrat or Republican we see none of that information. And as a result we would if we if we want to do ads we'd be very terrible at business cuz probably no one will buy our ads and we will show you kind of the wrong ads that you would have clicked on anyways. And so that's why our business model is really not compatible with ads and I would also say the ads business model is not compatible with privacy. privacy and this is a very clear distinction that we want to make. Now what if a parent were to going to say you know Andy this sounds good but we all know that sort of to be competitive or relevant today you know you're living in a Google world and so would it restrict my kid to like not give them access to all that Google tooling? Well it's actually giving them the choice. You know life is long and kids take a lot of time to grow right? Takes some 18 years before they come fully formed adults in in many cases. Once you put them in the Google ecosystem you can't take them out right? They're there for for good. There's nothing that you know prevents your child in 15 years from making a decision that actually they don't prefer the Proton ecosystem they want to be in Google instead but they can make that decision on their own fully conscious of the risk that that entails. So it's not either or. Uh it's just to give an additional option to children. So, if they decide that they want to live outside of Big Tech in the future and have an independent, uh let's say, uh future away from Big Tech, they have that possibility. Uh and I think this is a powerful option that you're giving your children for the future. Now, one more question about this. How uh important is this authentication option in terms of allowing these companies to build profiles of us? Uh for instance, ChatGPT or OpenAI is building a login with ChatGPT option. And I think that's like the utility of that might be even more than like, "Oh, it's easy to log in with Google because it, you know, you're in your Gmail." Uh maybe when you log in with ChatGPT, it can sort of fill out all your all those annoying forms for you. Uh what are you giving up when you log in with ChatGPT? Well, uh now, you don't need a login with ChatGPT to fill out forms. Uh you know, we have a Proton Pass product that fills out all your forms for you. So, so so, you know, it doesn't doesn't require AI to do that. I'll say, you know, this feature has been around for, uh you know, a long time. But, if you use like login with Google or login with ChatGPT, what you're doing is enabling a massive correlation engine. Uh it allows them to track your activity across multiple products and services and build a much stronger profile on who you are and, you know, your interests, what you're into. Uh and that is in fact, uh how Google has kind of gotten its, you know, tentacles everywhere, right? Uh by pushing things like login with Google. Uh and so, when you log in with Google on like a third-party service, you're essentially giving Google a full insight into your usage of this third-party service. Uh that's that's effectively what you are doing. Uh and this is why uh I say everybody that goes online today, you need to pick, you need to make a choice of who is going to own your identity. And today your options are basically meta, uh, Google, uh, you know, maybe Apple, uh, possibly Microsoft, but not really, right? Uh, and these options are all American, they're all big tech, uh, they're all spying on you, uh, you know, um, they're all, um, let's say, not the most responsible ethical, uh, tech businesses. Uh, and what I think Proton is trying to do is to say, if you want different option that has a business model that is aligned with, uh, you know, the best interests of, of users, and something that can guarantee your privacy through encryption, we want you to have option number five. Uh, you're not required to go through the store, but if you want to be completely outside of big tech, you now at least have this option. Okay. I didn't realize that when I'm logging with Google, Google, can Google see more than just the number of times I logged into that service? They can actually see what I'm doing within that service? >> Uh, it it depends on how the service is integrated. And it also depends on how often they're calling Google to authenticate. Uh, and and, you know, a lot of time these services are running Google Analytics, right? Uh, so all your visits are currently recorded. Uh, a lot of times these services are also pushed putting in the user activity, and pushing into Google Analytics as well from the back end. Uh, so, uh, yeah, I think a lot of times these services are from Google Ads as well. Um, so you so so it's a surprisingly large amount of information that that, you know, um, you get if you use log in with the Google option. Now, I mean, I think increasingly the question will be, all right, uh, if I want a privacy-forward alternative like Proton, um, it would be great to sort of go back to where we started this conversation for myself or my kids to have access to a leading foundational model or leading LLM. So you have Lumo, which is privacy-forward uh, AI LLM, but I'm also seeing that big tech is going to spend something like $700 billion this year alone on AI infrastructure. So, what's your plan to keep pace with them given that like [clears throat] you know, the one thing you could say about an ads business is it's high margin and they can just plow that money right back into building better AI models. Well, there's some people that say that spending number is just Nvidia paying Nvidia and you know, cycling revenue in a circle. And okay, maybe we have to leave it to the financial analyst to decide. But, interesting about AI that is different than the past cycles is if you look at today the most advanced proprietary models and you look at the most advanced open weight models and a lot of LLMs are open. The gap between the open models and the proprietary models is quite small. And actually decreasing over time. So, this is something that's very interesting about this cycle that is, you know, different. So, that also So, so what that basically means is even a company that isn't building a foundational model can get an AI product that is pretty comparable to, let's say, the top-of-the-line proprietary stuff. So, today if you use Lumo, for example, and you compare that with ChatGPT, yes, there's some things that doesn't do that ChatGPT does, but it is surprisingly competitive and in in fact quite close and the gap is closing, you know, as the open source and proprietary stuff are converging. So, this is a thing uh the first thing. But, the second thing that's interesting is there's also something called, you know, Moore's Law. And Moore's Law is basically that computing is doubling every 18 months and it's something that has surprisingly held true for a couple decades. Um And what this, in my view, will do in the long run is it will actually commoditize LLMs and commoditize AI. Uh, what I mean by this is today if you want to train a top-of-the-line you know our frontier model that might cost you a couple of billion dollars. Uh but in a couple of years that might be 50 million to you know 100 million. Uh and then a few more years on the line uh it could be one or two million. Uh and so what this really means is AI and language models in in large effect uh they're going to be commoditized. Uh and that's why I believe you know these billions that you see spending Uh uh yeah these numbers are huge right now but that is also going to exponentially uh go down with time. Uh and so I think the money is a factor yes but it's less of a factor as you go deeper and deeper into this revolution. Yeah and I think maybe something playing in your favor speaking of Nvidia's they're about to spend 26 billion on open-source models. Mhm. Right? So That's a good investment for Proton. Yes. Uh and and we benefit from that but it's also very smart for Nvidia to do that and the reason is Nvidia sells GPUs and they don't sell proprietary models. And they have great open models that everybody can use then everybody is going to buy GPUs. Uh so so so so so this is is also kind of an interesting trend where the most let's say the highest market cap company in the world which is Nvidia is strongly incentivized to develop strong open models which is going to make companies like Proton who are doing things privately first way uh have a massive advantage compared to the revolutions of tech that happened in the last 10 20 years. And this is and this is I think um quite uh novel. Uh and and this is for the world great because if you ask the average person like you know who uses AI everybody uses AI. But if you ask them like you know how many of them trust Sam Altman very very few actually trust Sam Altman. And this is I think also the opportunity for privacy companies because everybody needs AI but no one trusts it. And this is a problem that I think we can actually solve uniquely because of the dynamics that we just discussed. Definitely. Well, look, I'm definitely much more optimistic now having this last part of this conversation than I was when we came in thinking about how much of my personal information I've given over to these models. So, all right, Andy, and good news for those who do want that 15-year $1 email address reserved for their kids, we're going to put the link in the show notes where they can go to sign up. Yeah, I think it's by the way, I think another thing I want to mention about this is the dollar that we charge, you know, we really don't want to do it, but we have to do it in order to prevent abuse. But actually that dollar goes to the Proton Foundation. It goes to Proton's non-profit. And I think you know, so you also support a good cause. You support freedom of speech, you also support you know, democracy and freedom online with that $1 donation to the foundation. And you're giving your you know, child an option for the future. And I think that's a small price to pay to you know, give them a future outside of Big Tech. Definitely, Andy, I think one of my core takeaways from our discussion today is just it's important to have that option. It's important to be building that option. And so I really want to say thank you for coming back on the show and thanks for building the option for us. Yeah, thanks a lot and you're looking forward to see what the future will bring. And we want to build even more options for the future. So, I think it's >> [music] >> we're at a turning point in history where we are from a generation that didn't have options, but I think the next generation will have options. And this is I think something very important. All right. Well, Andy, always great to speak with you. Thanks for coming on. >> a lot. All right, everybody, thank you so much for watching. We will see you next time on Big Technology.