All right, everybody. Welcome back to All-In at Davos. We're here at the World Economic Forum. For some reason, they invited me and I'm here interviewing all the most important people in the All-In style, which is Full Contact.
We're always debating investment ideas on the pod. Well, Generated Assets on Public let you turn your ideas into an investable index with AI. You just enter a prompt like companies benefiting from power plant construction. Then their AI will build an index of curated stocks for you and back test it against the S&P 500. Then you can invest in it just like an ETF. Go to public.com and check out Generated Assets. Public investing for those who take it seriously.
I think most of us agree going into 2026, stable coins and AI, crypto having a huge resurgence and we were lucky enough to get two of my friends, Brian Armstrong from Coinbase, and Jeremy Miller, who's the CEO and co-founder of Circle. We've known each other for 30 years.
What a long strange trip it's been.
I know. It's great to be with you again.
Yeah. And this is not your first time at Davos. You've been here a couple times.
Yeah, I started coming with my last company, Brightcove, back in 2008, 2009. And very different time both for we it was a different time in the world. There was great financial crisis breaking out everywhere and that was an interesting backdrop and also notable I always reference that same week was when the first block of the Bitcoin blockchain was minted with the Chancellor on the brink bailout kind of embedded in the blockchain.
Yeah, it's fascinating. I think it's Game of Thrones. Chaos is a ladder.
Chaos is a ladder. Yeah, it's and for guys like us who've been through this I guess three times now. We went through it in the dot-com boom. I think we're too young to have we were you remember obviously that Black Friday in '87 but we were in college I think.
And so we had our great financial crisis and then we had COVID.
Yeah. Which was also pretty spicy.
That was a great disruptor. When you have those moments happen, I'm guessing for you and many of the founders I talk to, you just think this is the time to build.
Yeah, totally. I mean, constraints have a huge impact on what an entrepreneur does. And you know, even in the history of Circle, we've had extraordinary up and downs, some of which are endogenous shocks, exogenous shocks, all of this. And back to that 2009, that was that year actually I got my company profitable pre-COVID and then not long after it went public. And so we deal with what we're dealt with constraints. Yeah.
Yeah. Crisis makes for great art is I think the old expression. So let's talk about your journey with Circle. Stable coins obviously are top of mind because we had the Genius Act. My bestie David Sacks who's our crypto and AI czar for America who's here with me at Davos or I should say I'm here with him, he invited me to come. This technology is important. Why?
I mean look, when I got started working on this almost 13 years ago Bitcoin had emerged and it was from my perspective as an internet technologist I was thinking about wow this seems like a new infrastructure layer for the internet like a missing infrastructure layer of the internet. The internet had ways to represent data, media, audio, video, software but there was no notion of money on the internet and there was no protocol for money on the internet and it was very clear at the time that that was going to happen. And when we started, I wasn't convinced that everyone in the world is just going to use a new commodity money like Bitcoin. My view is that we needed a bridge. We needed to connect the existing fiat system to crypto and to these new networks and build what we called an HTTP for dollars on the internet. And that was the idea. And eventually that's called stable coins. Yeah.
Early on we talked about we're building fiat digital currency or fiat tokens or all this but stable coins stuck. But the basic idea is we now have a general purpose, general architecture form of money on the internet, digital dollars that can move just like everything else, and can transact peer-to-peer, which is super super powerful and provides a huge amount of value to people. And really importantly, because of the technology of blockchains, we actually have programmable money. And so we're right at the front edge of a kind of renaissance in how money is used in the world. We'll come back to AI, I'm sure, because it ties into that. But fundamentally, we just need a native way to have money on the internet. We need a very safe form of dollars on the internet, and that's what stable coins provide.
And variability and not knowing how much your dollar is worth when you open your wallet or your bank account was a blocker for consumers.
Yeah, I mean volatile cryptocurrencies, no one's going to buy a cup of coffee with a Bitcoin, etc. So, that's where the word stable coin came from. It's like, well, it's a coin, but it's stable. How do you actually achieve that is a whole different thing and that's where the model that we built, which is fully reserved with ultra safe assets and we have regulators look after it and auditors look after it and make sure it's all done the right way.
And I remember talking to you offline about this. It was the easier decision for most people in crypto to go to the Zug. The Zug. What is it? Zug.
The Zug here in Switzerland or just be based in Zurich or be based in maybe Dubai. And yeah, just YOLO it and don't worry about regulations and don't worry about audits. Don't worry about regulators. You made a different decision. You told me, 'No, I'm going to do this buttoned up and proper.' And why didn't you go for the quick, I'll just do this offshore and give up my United States citizenship.
Yes, we have many crypto people right now watching this show. Yeah. From San Quentin and [laughter]
Yeah. I mean, look, show's popular there. I actually it going back to even in the founding of the company when I looked at okay if we want to establish a new way to put what we think of as regular money on the internet and we want to actually build a new internet native financial system that the whole world uses, actually uses, businesses are going to use it and people are going to use it and we're going to do loans in it, we're going to do all these things with it. If you want it to work that way, well, you have to integrate with the existing system and you have to work with policy makers to figure that out. There's just no other way. So, I testified to the Senate in November of 2013 and if you read the testimony, I'm saying all the same stuff now as I did then.
And it was received slightly differently.
It really was. Yeah. I mean, well, I got a lot of hate very early on. Take yourself back to that moment in time where you're in the right. You're trying to explain how to do this correctly and you get a reaction that is different than what you maybe anticipated.
Well, I think with a lot of technologies on the internet, I was around I think we were both around in early stages of the internet, early stages of the web and there's a very kind of hard libertarian view and then there's the obviously the statist end of that. And I lean libertarian. But I think also, in the early days of the internet, it was like, well, if we want to allow companies to get on, well, we need to have regulation for ISPs and we need to be able to have secure transactions and SSL and well, then we're going to need people who are making decisions about who's a valid site and who's not a valid site. And these were really controversial things at the time that I think the more anarcho side of things was against. And now crypto itself, cryptography and crypto kind of was birthed out of that. And so these are people who wanted to be outside the system.
Wanted to be outside. Yeah. And so, and I think Circle has always kind of tried to find this middle way, which is open public networks, permissionless innovation, building on public blockchains, using open-source infrastructure, and really what you can do with something like USDC as a dollar and as a technology is far more open than the legacy payment systems and the legacy money systems. And so there is a very deep commitment to those fundamental internet ideas. But at the same time, if you want BlackRock to use it or you want the biggest tech companies in the world to use it or you just want someone to hold it in a digital wallet and be like, 'Yeah, this is actually a dollar and I can rely on it.' You end up needing to have this kind of structure around it for trust.
For trust. Yeah. Trust remains a key thing and obviously I think people talk about in code we trust is sort of what you'd kind of put around a easy for a developer to feel that way, easy for somebody who is extremely technical to feel that way right. But for a civilian, non-anarchist, right, yeah maybe you want to understand how the dollar is actually backed.
That's exactly right yeah so I think we just took that path and that was a harder path. It took more capital. It took more I had to hire my first executive was a general counsel and chief compliance officer.
Well you had to bring those in house I suppose because if you went to a law firm they would be like we don't know and don't take this risk.
Yeah. I mean one of the stories actually is that before Jim Breyer and General Catalyst would give me capital to when we were starting the company they were like we really need to know that if you do this it's not illegal like we're not going to have some kind of liability issue and so I personally on my own with my own money hired the top regulatory advisory firm in the world to go look closely at this, talk to people at the US Treasury Department, talk to others, really figure out can we do this and actually we could, there was a path to do it and that's because the US Treasury Department actually had given guidance about how to deal with virtual currency in the banking system. This was in March 2013 so super early. But yeah, I mean we needed to know is there a legitimate legal pathway to accomplishing what we want to accomplish? has always been important.
And as we've seen, there's things that are legal and then there are incumbents. Incumbents will use the legal system to try to stop innovators. You've also faced a little bit of that.
Yeah. I mean, I would say in the entire history of building this, it's been just there's been huge uphill battles with regulators and incumbents. And what's interesting is that I found over the years that if you come into a policy maker or a regulator and you say, 'Hey, there's this new technology, it can improve things in this way. We're trying to figure out how to deal with the risks. There are real risks. Let's not pretend there aren't risks and let's come up with ways to address that.' Actually people are pretty interested in talking. And so I think that's there. But generally when you go to the incumbents and say, 'Hey, we want to work with you or integrate this because we kind of need to work with you to make the whole thing happen' a lot more skepticism and constraint there. And so that's changed now. I mean, we're in a very different world now, but now it's gone from a threat to an opportunity.
But there is still a little reticence to stable coins and crypto by I think some of the major banks and the big players because they're maybe they're my perception is they're a little concerned that you're too good at what you do and you might be too far ahead and if you weren't so dextrous at doing this and capable and they started, 5 years ago, they might feel differently about it. Is my intuition close to your assessment or am I wrong?
I think that there's some truth in that for sure. But it's interesting like it's both you and I have this kind of juxtaposition against these other eras of the internet and when digital media happened and it was like hey you could stream media or you could put up content on the web and you could do all this there were all these digital media startups and the media companies were like well we can do that too and okay our business model might have to change etc. and digital advertising, okay, well, you can transform advertising into this. Oh, it's more targeted. It's all this stuff. But then obviously were companies that kind of executed fundamentally different technical execution, software execution. They built completely different types of utilities with completely different unit economics that kind of turned the product user experience and the economics upside down. and those we know those companies.
Craigslist to classified, Google ad network, New York Times marketplace, and many many more like that.
And so I think we're in a similar place right now where a lot of media companies absolutely and communications companies as well and enterprise software companies there's big buckets here retailers sort of said okay this is a new paradigm it's better for customers it's I can deliver a better product like the economics are better we just have to do this and we're going to innovators dilemma blah blah blah and they make that transition and it takes like 5 to 10 years.
It takes them three times longer than probably the innovator like Walmart and Target actually do an exceptional job now. The United Airlines app is not terrible, right? I mean, it's not Uber, but and this is the difference between a software-driven company and a non like we're a software company like we fundamentally of course we're going to go faster, of course we're going to understand UX. So the banks now my understanding is some of them are threatened by the concept of a stable coin having the ability to generate points incentives or interest essentially and that's kind of the sticking point. The Genius Act says hey you can have rewards but you can't have interest and that was I guess where this was able to get through the legislative process. Explain that to the audience.
Yeah. So, a couple things. So if you go back a few years when stable coins sort of emerged, the sort of all the biggest regulators, bank regulators around the world got together, what's called the Financial Stability Board and said we have to have regulations around this like this can't just go crazy. That was also when Libra was coming out and all this. It didn't come out, but you tried to come out.
Meta tried to do their own project and they gave up on that project based on my insider information because the government felt they had too much power already with their platform and Zuckerberg said I have enough heat on this company already. I don't want to have another group of people thinking I have too much power and he shut it down. That's what I was told by insiders.
It's interesting is you know one is that your understanding as well? I have a variety of insights about it, but I think we're kind of a neutral company, right? We're not a giant big tech, etc. But kind of coming back to your question, regulators got together and said, 'Hey, there is this payment system innovation. It is this thing called stable coins.' And they all kind of said, 'Here's some recommendations for how you could regulate this.' And actually all around the world that happened. So it happened first in Japan. there's stablecoin laws and then it happened in Europe years ago actually stable coin laws and then in the UAE and Hong Kong and then in the US with the Genius Act and in all of these stable coins are designed as a cash-like instrument a payment instrument they're designed as this form of money to be used in the payment system and that's really at the heart of it and that's how these laws have been written and it's pretty consistent around in the world. So, Genius Act does that, but it also I think provides so as a result under the Genius Act Circle as a payment stable coin issuer, we're prohibited from paying interest directly to stable coin holders. It's the same thing in Europe. We're regulated in Europe. Same rule. Same thing in Japan all these markets is the same. But at the same time, we're building a business. we do generate revenue and we work with lots of different types of platforms and markets and distributors and brokerages. We work with everyone from Robinhood to Revolut to Coinbase to Visa and lots of companies. And so those companies, stable coins are really important too because they are how people hold money, how people make payments, how people trade, how people do a lot on those platforms. And so they want to be able to have the ability to pay rewards or have loyalty programs or incentives and other things. And if they're making money from their relationship with us, they want to be able to do that. And that's what the Genius Act captured. And I think really I think it's good. I think it's a good model.
So you're happy with it as is?
We think it's a very solid model. And I think really there's some ongoing discussion about well is there a specific set of things that qualify for those kinds of loyalty and rewards and other things and how prescriptive should the law be? And it's kind of getting relitigated a little bit.
The banks, my understanding, maybe want to try to kill it. Maybe some of them who feel the most threatened. Is that the back channel here at Davos?
I mean, I think banks see stable coins as both a threat and an opportunity, right? You know, I can say we're having more engagement with more banks in the world than we've ever had before.
And so and major banks, global banks, regional banks, banks all over the world who want to integrate this and use it in payments. they want to use it in how capital markets work. Think about I need to post collateral on an exchange to trade and I'm trading different assets. Stable coins are a better way to do that. And so whether it's in capital markets in trading in wealth management in payments there's uses for this and so definitely banks see a lot of.
Because it's faster and cheaper. It's instant and it's free essentially.
Yeah. I mean if you talk to a global money center bank you can imagine one of those like.
Something like that or a Citi or one of these banks and if you ask them how many different payment networks are you integrated into as a bank and they'll tell you like over 200 they integrate to over 200 different payment networks.
And it's all over the planet.
It's all over the planet.
And it's all different standards I would say.
All different standards all different systems etc and so if you go to them and say well, stablecoin networks are like a new payment network. They're like, I get it. Like, this is, oh, it's on the internet. Oh, it has the attributes of how the internet works. We understand that.
And this will be your fastest, cheapest, and most trackable version.
That's right. And so, we actually we have, there's a global systemically important bank we work with who's actually moving their own money between their own global branches using USDC because it's faster than going through the correspondent.
They trust it more. Yeah, in those cases. So bottom line is there's a lot of opportunity for banks. Bank it is not this is not a black or white thing. I think we'll figure out this rewards thing. I'm very sure.
Yeah. I mean AMX figured it out and video games figured it out. There was a little rattling even in those cases where they were like hey is this a currency because people are trading it on eBay trading my sword from World of Warcraft or whatever these nerds are up to. And they were like, 'Huh, is that like you have your own currency?' It's like, they're playing video games. Who cares? Like, can we just move on? There's more important things to do in the world than sweat.
Pretending it's a currency. You're up against an offshore platform that has I think maybe three times the amount of assets under management. Tether. Who's counting?
Three, four, five times. But you have been catching up pretty quick. Yeah.
Yeah. USDC amongst the largest stable coins has been growing faster.
You're number two for two years straight. Yeah. We're the second largest.
And we are by far the largest regulated stable coin network, right? And our, kind of growth in the amount of transactions happening with USDC is growing quite a bit faster as well. So, yeah, we feel really good. I think our view has been to the earlier question if this is going to be part of the actual economic system. If households and firms and corporations and financial institutions are going to use this and depend on this, whether you're a remote worker who's a software engineer working in Pakistan or you're a small business that's importing products from Vietnam that's in Brazil or you're a hedge fund in the US that's trading derivatives, you're going to want to know that you have something that is safe, fully reserved, audited, compliant, liquid, available in and out of the global financial system. And when we think about that opportunity, that's an enormous opportunity. The TAM of legal electronic money today is about $120 trillion and growing because of monetary easing and things like that. But of that, there's about $60 trillion, which is physical cash and non-interest bearing demand deposits. So kind of working capital money sitting out there, right? So there's a huge amount of value stored and used as a store of value as a payment system money and stable coins can grow into that and there's going to be an infrastructure for that that's well integrated.
And the business is there's a float on this large base of capital that gets to underwrite your business if interest rates are high it's boom times Tether's making over $10 billion a year on their float you're making I'm assuming billions of dollars or low billions of dollars on your float. I'm not sure if that's public knowledge.
We're a public company. You can look it up.
Yeah. So that I mean I'm assuming it's billions. Yeah. Well, I'm not sure if you break it down by that.
We talk about reserve income.
And it's billions, it's a significant amount.
And so but if we go into a zero interest rate phenomenon, this could be challenging for the business or not.
Well, a couple things. I mean, if you actually look at what's happened, when interest rates were actually very low, we saw 100% year-over-year growth, two years straight. So, growth was off the charts. When interest rates actually started to rise, we actually saw declines in circulation.
Well, it's really the price of money, the opportunity cost of money. So, interest rates set a kind of cost to holding money. And so the incentive matters.
Yeah, the incentive matters.
And so what's interesting is that if you actually look at from December 2023 when basically the forward curve which was sort of the short-term kind of price of money in a sense the forward curve kind of being the market's view of the short-term price of money when that started to fall because the expectation of interest rates came USDC started to grow and actually from the peak of whatever was five and a quarter five and a half down to where we are which is three and a half for whatever it is. Now, that's I don't know exact percentage, let's call it 35-40% decline in the interest rate. We've had a multi-hundred percent increase in the amount of USDC in circulation. And so there's an inverse correlation there. And so I've and I've said this publicly, I've said this publicly in the media, I've said this many many times. I have wanted when interest rates are really high, my view is we really need interest rates to come down. We really need them to come down because that will help us grow. That will put more velocity of money.
The pie gets bigger. Yeah.
Pie gets bigger and which also impacts adoption.
It impacts adoption as well. So more people are investing capital in technology to transform their businesses to grow.
Tech on right and so that is a catalyst.
And so my view has been that is really important and so I think our view is there is some conceptual neutral interest rate and given the persistence of inflation of around 2.5% or whatever it is some people argue is the neutral rate should the neutral rate and the kind of baseline of inflation be 2.5 or 2.8 or.
You know where the 2% came from? I talked about it on a previous All-In episode. I don't know if.
Yeah. There was a politician in New Zealand was talking to their central bank and he's like what should it be? And he's like I think two and he's like yeah how'd you come up with the number? It's like my instinct.
4% unemployment 2% interest rates. Both of those are going to be challenged. both of those are going to be challenged because of very significant technological and macro forces.
Yeah. Job displacement, globalization have had impacts on this depending on the country. The unemployment rate and also how many people choose to be employed. I mean when you have a society that is doing as well as ours is or the European Union is some people just choose to not work. labor participation now is 61 or 62% and then when you and I were coming up it had peaked at 68-69% so it was like.
It's really interesting to think about that.
Tether as a competitor would not be able to operate in the US lots of reports I'll be I'm going to be judicious here claims reports concerns from politicians to the press and everywhere in between and regulators that they're doing some in the dark areas that maybe you wouldn't want to touch in terms of the dark web, dark economy. But rumor is they're going to do a US version of Tether and maybe try to go legit and audited. How do you think about them as a competitor given that they've been banned in multiple geos and there was some very big concerns that they might have a run because they weren't dollar for dollar backed and people had all kinds of claims about hey maybe they have Chinese paper and all these different concerns and they wouldn't do an attestation or they would do an attestation with a bank and somewhere in the BVI that nobody's heard of and not a big KPMG maybe. A lot about this kind of I was not a Tether truther, but I did go down the rabbit hole because I reported on it on the pod and talked about it. But they are a significant player. I'm wondering how you think about them coming to the US and becoming more legit and maybe trying to catch up with you in that.
I mean, look, the beauty of having the Genius Act is that it creates a level playing field. It defines a federal law that you have to come in not just be audited, but you're regulated by if you're large by the national bank regulator, the OCC, which is a very serious regulator, a serious credential regulator. And we've conditionally we've received conditional approval for something called First National Digital Currency Bank which is a national trust bank that we're setting up to be part of how we operate USDC as well. But again level playing field if you want to comply and build a product under that regulatory framework you can do that and I think many people will and this is like when net neutrality happened and common carrier rules came like anybody can get into providing data services to the internet anyone can build these things so.
That's great and so free market clear rules love it so my view is there's absolutely going to be more competition and my view is also that the structure of this market stable coins are network businesses meaning they actually exist as platforms and utilities on the internet like the USDC and our stablecoin network is literally the software protocols but it's also the tens of thousands of applications that have integrated to the APIs and every time an app integrates Cash App just said, 'Hey, we're adding USDC support.' That's great. Now, all those Coinbase has it, Revolut has it, banks have it, Visa's using it. Every time someone adds it, Stripe merchants, Shopify merchants can use it. Every time someone adds that, it adds utility to the network and it adds network effects. And then the next developer who comes along and says, 'Hey, I want to build an app that uses digital dollars, etc. Which one should I use?' Well, I have interoperability with all this stuff. Great. So you have these network effects that are really key and then you also have, what I call liquidity network effects, which is the ability for one to easily get it and use it within banking systems all around the world. So we've built out this incredible liquidity network by being regulated in Singapore, in the UAE, in the European Union.
And this isn't an easy thing to do. I mean, this is a huge build.
It's a huge build. And now it's at a point where.
It is defensibility and we have pipes in a sense where some of the biggest capital markets participants in the world companies like BlackRock can enable a tokenized fund that enables USDC to come in and out of it and their institutional participants know that'll work right. So, I think it's a long-winded answer, but the answer is essentially we feel like we've built a great platform and a great network and we have really strong network effects and they're accelerating and we've been publicly audited as a corporation.
Yeah. As a New York Stock Exchange governed, publicly listed SEC supervised company now for a period of time, but more broadly for a long time. And when companies are going to choose what they're going to build on and what they're going to use, they're going to look at all that.
Yeah. The footprint matters.
It does. And so, the marginal value of a net new dollar stable coin coming into the market is essentially zero.
Yeah. You need, all these things are needed.
And what do you think about this concept that everybody's going to have their own stable coin? There'll be an Amazon one. Obviously, PayPal's added one. Do you think that's realistic or do you think people are going to be like, 'That's just I might as well just use Circle.'
Yeah. I don't see that. I think this is a lot like other internet platform markets like everyone doesn't need their own data center. Everyone didn't need their own vertical search engines. You know, there are a lot of things that people thought everyone didn't need to build their own video platform. Yeah. And so internet scale utilities achieve network effects, they achieve unit economics, they achieve developer flywheels, like all these things and they tend to lower their prices over time. it tends to get more and more economical and more and more capable and so there's a lot one would need to do to do that and I think even the biggest banks who have pretty big technology budgets may reach that same conclusion.
What worries you now? What's 2026 forward looking like? What keeps you up at night? The opportunity is obvious. I think you've explained it perfectly.
Yeah, I mean look, I think we're at an interesting place where we have a big piece of the regulatory clarity done. There's more pieces that are needed. So there's work related to markets, related to how digital tokens work. That's actually really really important because we believe in tokenization as a phenomenon, smart contracts as a way that people are going to intermediate things on the internet. So there's more that's needed there. And I think now, I think one of the things that concerns me is the changing geopolitical geoeconomic landscape, right? Which is there's lots of different points of view that are being taken on where economic alignment is, etc. And as a global company that is building technology that we want to make available globally, that introduces new complexity, right?
Yes. There we are entering an era of new alliances and national champions.
And you're an American company and America first and so yeah, you might be looked at differently based upon how the American enterprise is generally done.
Yeah, that is very true. our headquarters is Freedom Tower in New York City. and at the same time the technology of these networks there's an opportunity actually to develop versions of this that are more geopolitically neutral, geoeconomically neutral so that if you're from India or you're from Brazil or you're from Southeast Asia or you're from the Middle East you can build and these are high growth markets that you're comfortable building on this with a variety of different nexuses of economic relationships. And so that's something we think about actually.
Yeah. Like a euro-based stable coin.
We have the largest Euro based stable coin.
How many currencies have you done that with so far?
We've really only focused on dollars and euros. And.
Well, I think those are two widely adopted currencies for both as a store of value and as freely floatable, circulated, tradable currencies. And we have a big commitment to the European market. And so it's also part of just being committed to that market. We think it's a big opportunity. And then for everything else, we've really built a technology stack for other stable coin issuers from other markets now that there's laws around the world.
Ah, so if they want to do their national stable coin centralized, are you doing that with Bermuda or some sort of partnership?
The Bermuda partnership is related to helping them kind of have digital dollars be and have economic activity, commerce activity, treasury activity all happen on chain.
Got it. So that's infrastructure play, not I don't know what their dollar is.
Well, they're a dollar. It's a dollar-backed currency. They're already their own stable coin. But in other places, in the Philippines or in Mexico or in Brazil or in Japan and in other markets or Australia, Korea, there are stable coins coming and we want to make sure that we can provide technology so that they can do all the same things.
To the government or to banks.
Well, more to the private sector because stable coins are private sector innovated, but these companies all need to be regulated by the government. They need to be following the same.
Easier for you to partner with a major partner in those markets.
Yeah. And we want to see that grow and flourish. You know, this idea of an internet financial system where, all the economies of the world can be onchain and all of the contracts and financial contracts and markets and capital formation, lending, everything.
Can happen on chain. It's obvious that small businesses and people who are very and poker players, people who are concerned about their fees
will and and you you know are focused on that are being driven to stable coins. Like they they get it. They're like, 'Yeah, I'm doing my design job for you know, I'm a designer in the Philippines working for somebody in India.' They're starting to figure out, hey, I don't need to pay these. And then consumers are starting to figure it out.
Yeah, like me, a poker player, you're sending a lot of wires. Every year I send tons of wires back and forth for different poker games.
Well, now you know Polymarket and Kalshi both you fund your account with USDC, right?
And well that's because that audience understands because they're always looking for edge speed. Speed and but they're looking for edge in a poker game like you as if you were a novice and I was an expert. If I was Jason and you were Phil Helmuth, whatever it is, you know, it's got a bigger edge on him. It's an inside joke. Anyway, point is like you might have a 1% or 2% advantage, but it compounds after 100 poker games to be a bigger number.
It does. This is why the rake matters and you know gamblers and people like to wager things.
You know who the biggest users of USDC are?
Yeah, that's what I was kind of getting at. It's actually it is a form of gamblers which is the biggest electronic markets firms in the world.
Got it. The guys who need every edge in terms of milliseconds and cost and capital efficiency to be able to move capital in markets because at the end of the day they are algorithmically trying to figure out the very next best thing to do. And if they have dollars that operate with the physics of the internet and the cost efficiency of data, they have an edge. And so they love it.
They love it. And then the people who are sending money home to their family, they love it too.
Yeah, because they're like wait I'm paying Western Union 12% of this this is crazy.
Right, they can have a wallet and directly send to their counterparty and it's like wait a minute this literally arrived and there's proliferation of these digital wallets now that also can where you can hold it in USDC but spend it through Apple Pay, that are being issued all around the world too. So people can store value in digital dollars, use it through payment terminals, and that's pretty cool.
TurboTax thesis there because that seemed like there's something underlying that that I wasn't getting to. I mean, it's obvious that like okay, there's money flowing through there, but what's the big picture there?
Well, Intuit is an amazing company. Obviously, they've been around for a long time. They have franchises that are huge. Credit Karma is a huge franchise, actually huge. They're growing what they provide to people through that franchise. QuickBooks, obviously, we all know if you run a small business, you use QuickBooks. And actually QuickBooks, you know, they invoice people through QuickBooks and it's actually trillions of dollars of transactions a year that are invoiced through QuickBooks. And if you can invoice people and settle in USDC, that's better than ACH and it's faster. And you can imagine a world where small businesses can gain an advantage from something like that reconciliation. If you're a TurboTax user and you file your taxes and you get a refund, imagine being able to get your refund instantly. So I think there's a lot of cool stuff and Intuit is a very innovative company. It's a tech company that is in these financial technology adjacencies, and so we're super excited about the collaboration that we have.
I don't want to say there's a loser in this, but there are people who are going to be challenged by it. How does American Express and Bank of America look at Circle and stable coins in your mind?
Because they must call at some point or do they not? I think all these companies have an opportunity to use this technology, right? They have an opportunity to use this technology to improve how they provide payment utility. They have an opportunity to use it if their customers want to interact with the whole on-chain economy and on-chain investments and what's happening there from a wealth management perspective. I eventually think that there's going to be more credit products that are actually built using stablecoins.
How would that work? And what's the advantage?
Well, so already today, there have been trillions of dollars of loans made in stablecoins through DeFi protocols. Yes. So you take your USDC and you effectively deposit it into a protocol, and on the other side of that protocol is a borrower, and you're paid an interest rate for the amount of time.
You become the house, you become the bank.
Well, the individual does actually. I mean these are as someone years ago said, these are like self-driving banks. Meaning it's software machines where the risk management, the collateral management, the liquidation, all of the liquidity that's there is all just smart contract machines that we can all observe in real time, perfectly auditable, transparent. You don't have that with a bank. So you have a really interesting thing there. And right now a lot of that is like lending to people who are borrowing to do things like investing, like margin-type.
Like margin loans. If I own a million dollars in Bitcoin and I want to spend 100,000 of it but I don't want to liquidate any Bitcoin, somebody else can make some yield.
And you can borrow USDC against your Bitcoin and so on. But I think this is the early stage of that. And so I think our view is that if you have these digital cash things like USDC, you can create lending protocols that are lending for you know, I want to hire a new employee or I need some new equipment for my kitchen in my restaurant or factoring, equipment leasing.
All these forms of lending right can be done, and the risk that is taken, there's real risk there, can be underwritten, and the insurance on the risk failing can be priced. So you can build credit markets entirely in software, entirely with software machines coming into play.
AI plus these sort of smart contract machines and stablecoins, I think creates a really interesting little cauldron for credit market innovation.
Yeah. And I have a glean in my eye, which is imagine a credit market that worked like AdWords. Imagine something that was that efficient and could clear and settle credit decisions at the speed of which an auction happens for attention. Yeah. And so those kinds of things will become possible, and that probably would be pretty good for people who are on the other side of credit need.
Yeah. If you think about also monetary velocity and the economy, you can get more things moving.
Exactly. More jobs created, more chances, more swings at bat, more shots on goal.
So money velocity increases. And I think these kinds of credit intermediation models powered by AI and blockchain networks can actually further increase money velocity. And you know, we have a mission statement to increase global economic prosperity through the frictionless exchange of value. That is literally our mission statement. And people a lot of times they think exchange of value means like making a payment, like no. Value exchange is the sort of time value of money transformation. And it's sort of, I have value I don't need right now. You have a need for that value, and I'm going to use time to transform that and then create new things from it. And that's where actual growth comes from. So that's what we want to see.
One of the interesting things since we're here talking politics at Davos, you and I are talking tech, but politics is you know one of the big topics here. So we'll sort of end on that a little bit and get your ideas and your thoughts on it. There's a movement towards socialism in New York City, my hometown. It's kind of heartbreaking for me to watch. California is seizing the billionaire's assets, literally asking them to write down and audit everything they own, put a value on the painting or the piece of jewelry they bought for their spouse, and we'll just take 5% of it once. And they lost 200 billionaires and a trillion dollars worth of, and they don't seem to have any problem with it. As a technologist, a builder, a capitalist, as long as I've known you, what do you think about this moment in time us Gen Xers are living in watching this socialist movement?
It's interesting. I just moved to New York City. Yeah. Just in time.
Just in time. Yeah. By the way, that's costing you I think you're going to have to pay 2% more a year, but I don't think you're an income guy. I think you're more of a cap gains guy.
Yeah. Look, here's what I would say. And actually I was with Dario from Anthropic a couple times over the past couple days, and he obviously talks a lot about the economic disruption, labor market disruption, what's going to happen with AI. He's not a doomer, but he has deep concerns. And I also do, which is when I look at the very positive thing, which is that I think we're going to see an acceleration of GDP growth. Pick your time frame 3, 5, 10, 15 years, we're going to see an extraordinary period of economic growth acceleration, even more than the internet and PCs.
More than the internet and PCs.
This is a very, very big transformation, but where most of the labor is actually executed by AI machines.
And the capital accrues to the capital owners.
And that's just kind of there in front of us.
Yeah. It's the fundamental nature of capitalism.
Fundamental nature. And I think that is just going to challenge us in ways that we haven't dealt with.
In a long time, since other periods, labor was necessary for capital.
Yeah. The industrial revolution was mechanized human labor, right? Yeah.
This is, you know, mechanized AI.
And by the way, the CEO job seems like ripe for AI doing a better job than we would do at times, right? Like I don't know if you've...
I use AI all the time in my...
I don't know if they realize what they're interacting with. No, but I mean, I literally when I was talking to Brian from Coinbase, our friend, he said he's got all of his data into a proprietary LLM internally. We said this on the program. I'm not speaking out of school. And that he asks it, 'What do I need to know about my organization?' It's like, 'Oh, there's non-consensus about this important issue in the corner over here.' And he's like, 'I didn't know that.'
Like, whoa, is that a coach or is that like a parallel?
Like another CEO? Was it a duplicate CEO? Did you clone yourself? How do you think about that?
I mean, I think about it all the time. I'm pushing very hard to encourage that we are all inside of Circle doing more with AI. We're making sure obviously with safety and compliance in mind because we're a regulated company.
Yeah. So there's a lot of work. We invest a lot in that, but nonetheless, creating the avenues so that that awareness and understanding can be there and that as leaders we can turn to that. I put a Slack post in a Slack channel recently that sort of said if you want to become a great manager at Circle, probably the one of the best things that you could learn how to do is manage AI.
Yes. And so invest in how you manage AI and AI agents actually that's how you might actually become a better manager and leader through that. And I mean that. It is one of the if you think about the tasks of these product managers who would run the standups or you know that your mid-level manager historically it's to keep tabs on a group of people and their targets and their time frames and who's excelling and who's got blockers. AI is kind of better suited for that.
And so the a great middle manager, which is a derogatory term, but let's just say you use the word manager. A great manager should be able to manage 10 groups using say Claude's co-work. I don't know if you've been playing with it yet?
I mean, it's out a week and I'm like I'm sitting there like...
We haven't deployed that inside the company yet, but I've outside looked at it and used it.
What impresses you about that?
I mean, look, the ability to take tasks that I have to do and just teach them and then it just can do them is extraordinary. And I'm trying to get this deployed in my household too. But it's really amazing.
Interesting way to do it is uh because I went down this rabbit hole like estate management stuff like when you have to deal with this when you get a second home or however many you wind up having.
Notion is I think the best solution right now because Notion put AI into it. Do you use Slack personally too for your domestic affairs?
Uh, you don't have a domestic?
I don't have a domestic Slack.
Okay. I have a domestic Slack and a domestic Notion. So I took the stack from the office and I just did it at home. And what that did for us was the people who bridge you know the family office and this you know can kind of they know the same tool stack. Then Notion has been investing in their AI so much that when you do a query like an AI query of your Notion and you're like 'hey this house with this HVAC you know this issue' it's like answer! You're like 'oh my god that would have taken like three phone calls and an hour.' And I think it's going to change everything about how we work.
Is what what what else did Dario think? I think I might have cut you off before you...
Yeah. I mean, look, I think the big picture there was just like whatever your politics are, whatever you think about taxes, whatever you think about the social contract, all that's I think in the coming years going to kind of have to get thought about again. Yeah. In different ways. And I don't think it's going to be the straightforward answers that we've turned to. No, it can't be. And so that's why I, you know, my answer is like yeah, there's things California, New York, etc. are doing, but I think when we look at that in comparison to the broader political economy that we're going to have in front of us, those will seem modest in comparison to what we've got to think.
By the way, we fixed it in startup land long ago. It's called stock options. And if you look at the Trump accounts, which Invest America, Michael Dell did, and Brad Gersonner and a number of our friends, if the 42% of Americans who don't own an equity and you know, are not part of that, let's just say it's 50 because some of them own them kind of passively. Half the country doesn't own equities. If everybody had equities in a superannuation fund like they have in Australia and they're watching them go up, then you're not like, well, screw those guys at Amazon and Apple and Google.
Being stakeholders in the success of this technological transformation is critical.
Yeah, that would I think solve so much of the problem. Jeremy, I could talk to you for hours. You and I have talked for hours in our lifetime about all these paradigms. Continued success and it's just a pleasure to spend time with you. It's always like one of the great discussions I have.
I can't believe it. George Kurtz is here. He's the CEO and co-founder of CrowdStrike.
This is your second Davos.
Why are you at Davos? Obviously, everybody knows CrowdStrike. Yeah, we'll get into it. Why are you spending the week at Davos?
Well, I have all the world leaders here and all the best company CEOs are here. So really, it would take me a year of activity and flying around the globe to meet all the people that I need to meet and it's just a fantastic opportunity to meet customers and get business done.
It's this tiny little ski town.
It's not fancy at all. This is in Aspen and everybody's within five blocks once you get through security.
Speaking of security, AI has had or is going to have I want to get your opinion on this, a profound impact on defense and also offense, black hat, white hat, everything in between. What is the attack vector that AI is most effective at sitting here in 2026?
Well, there's a couple of different areas. One is if we think about the adversaries, I kind of create a pyramid where you've got nation state at the top, very sophisticated, but less of them. The middle band of the pyramid is e-crime. More of them, a little less sophisticated than the top. And then the bottom is just activism. So essentially, you're minting new adversaries because you don't have to have all the knowledge that you had to have in the past. You can ask any number of LLMs and you can get answers back. So what's happened is the attack timeline has been compressed. You can automate all of these sort of attacks and you can do it with a level of sophistication that looks like a nation state. That's one of the greatest areas of exposure is the speed and the sophistication has dramatically increased.
Because the answers have been compiled. Yeah.
The scenarios can be run even by somebody who is a level six hacker out of 10. I'm just making a number up here and they can become an eight. So just like a developer using a co-pilot goes from being an average developer to an above average. The same thing is happening with hackers.
Absolutely. And one of the things that we're seeing is autonomous malware. So most people are probably familiar with malware runs on your computer. Typically people would use antivirus to protect against that malware. But what we're seeing now is prompt-only autonomous malware, meaning I can drop the prompt on your computer by a variety of means and then it will autonomously interact with an LLM and it will give a unique fingerprint every time it runs. So your computer is different than the next guy's. You have different data and it will begin to prompt its way to getting to what it needs without ever phoning home to someone controlling it.
Oh wow. So just to unpack that and translate it. I did a little hacking in my day in the 80s. Okay. A little phone phreaking mainly. We'll get into it offline. All right. I think I had a Haze 2400 guilty charge. I started with the Ventel 300 baud. I think we're of the same.
It was a tank. It literally had such a great form factor that's never been repeated. This great like rectangle, but we're old.
We are old. People are now saying, 'Hey, I'm going to use the Comet browser by Perplexity. Chat GPT's got a browser. Claude has an extension.' And it's really wonderful to say, 'Hey, get all my emails from LinkedIn and then put that into a database and add it to my CRM.' Wonderful. You watch it work. Get me the cheapest flight. The same thing can be used by an adversary to get on your computer. And one of the great ways is detection in security. You're in the business of detecting the attacker. But what you've just informed me of which kind of blew my mind even though it's completely obvious is the LLM can say, 'I'm going to do a thousand attacks today and I'm going to make each one different and unique.'
And unique. Absolutely. They're unique.
They're not actually code because they're prompts, right? And traditionally malware would phone home. They would be there would be like a tether, right? So you'd always be able to see the signature of something phoning home. Now these can run autonomously without ever phoning home.
Like a sleeper agent. You drop them into America 20 years later they start pursuing, but they never have to return to base.
How does one counter that?
Well, you need AI to counter it.
And I mean, that's been a big part of the success at CrowdStrike. When I started the company, I founded in 2011, it was using, we'll say AI, but machine learning at the time to detect malware that's never been seen before. Now, obviously, that's evolved into Gen AI and really leveraging basically the large data set that we have that we've amassed over 14 years to train our own models to be able to counteract with the speed that you need what the adversary is doing. So, we took, if you think about it, there's only a few ways to break into a bank. I'll use this analogy, right? You can drive there, you can walk there, you can use a gun, you can blow the safe, but at the end of the day, you got to get the money and you got to get out. Yeah. And these, it's the same way in the computer world. There's certain we call indicators of attack. We've trained our models to look for these. So, it doesn't matter kind of what you look like or what car you're driving. We can identify that. A big part of our success has been the AI we've built over the years with this tremendous data set.
Yeah. So, you might be able to get into the bank, but at some point you got to leave with the diamonds and you got to get to the safe. Exactly.
That's where you can catch him.
Uh you mentioned state actors and you mentioned for profit. When we look at a country like North Korea, they need revenue.
Yep.
Hacking is pretty great revenue source. They were even using AI, my understanding, and I'm sure you've been following this, to get developer jobs in the United States, convince Americans to put laptops in their homes so that they could remote work. Talk to me about that attack vector actually infiltrating companies as employees because of this remote work nonsense.
Well, we were the one of the first to ever find that. So, we were actually going through developing some new AI algorithms. Yeah. And we saw something that was called signal, which basically strips the noise from the signal. And we saw the signal and we said this is really weird and we investigated it and we said okay this looks like somebody using remote tools to you know what's going on here. And we went further and then we said, 'We think it's North Korea and we think that it's an employee who the company thinks is an employee.' So when the R&D team came back to me, they said, 'Okay, we want to notify our customers and we want to say we think we found this, but we have to tell them that the employee that they think is an employee isn't really their employee.'
Yeah. The phone call's coming from inside the building.
Exactly. So for us, it was, you know, when you have a piece of malware, it's like, 'Okay, that's bad. You can detect it.' But when you have to tell a company that their employee may not be their employee, there's, you know, you have to be gentle.
It's a little gentle. Like you're letting them know just how incompetent they are when it comes to security. I said it, not you.
Okay. Well, it's hard to find these guys. But in any event, we found 40 of them. We were right. 40 of them. That's for the first run. We found hundreds now over.
Now, they were doing it, my understanding, for the high paying salary to fund stuff in North Korea. Is that correct or not? They were doing it to get trade secrets.
They were doing it to buy access. Why break in when you can just log in.
You know, so it's a lot easier to get somebody hired. So true story. So we actually say we think this employee is not an employee. Company investigated. They investigate it and they go, 'Yeah, you were right.' I said, 'Well, tell me the story.' And they said, 'Well, we went to this person's boss and we said, we don't think that's a real person.' And they went through all the reasons why. Finally, they said, 'It's, you know, we think it's a North Korean.' And the guy, his boss said, 'Well, do we have to get rid of him because he did such good work?'
Oh my lord. So the value of a top tier developer is so high that you'll take us by.
He's like, 'Ah, you know, we kind of need him to ship.'
Yeah. He was doing such good work.
He was he was shipping. This is one of our best performers. True story. What now? I mean, remote work has opened up so many factors. You believe in remote work. Do you support it in the company?
We do. Yeah. We started the company in a remote first.
But this opens up all these attack vectors. What's best practice?
Well, even even without remote? Well, best practice really and what we see companies doing is we'll take a simple one. Meet whoever you hire.
Shocking. But you know with co-working and remote work and things of that nature. There's a lot of people that get hired that people never met and you know you never see them on the screen and but the work would get done. You know emails would get written and then you know these things happen. So what we're seeing now is the best companies are actually embedding a security person in the HR group where they're prefiltering all these resumes because the resumes are AI generated. The LinkedIn profiles are AI generated. So you want to catch them up front rather than in the interview process.
Crazy idea. Hey, for your final interview, can we fly in or hey week one you're going to be at HQ.
That's it. You solve the whole problem because then they don't take the job. They move on to the next person.
Okay. So we're looking at state actors. Russia still number one state actor.
China, Russia. I mean, it depends like if you you know, pure intelligence, pure ability, Russia.
Russia. Yeah. Why? Why are they so good?
I mean, look, the Cold War you could say is over, but you have really smart people that went to work in various areas, right? Obviously from a nation state, but also from an e-crime perspective. They're smart. They know what they're doing. Typically, they're what we call low and slow. The Chinese over the years have been a little bit noisy, kind of a smash and grab. They've gotten better.
Not subtle, but much better now. But the Russians are very targeted. They're patient and they're not always gathering IP information to steal it, but they're gathering intelligence for reconnaissance for nation state activities and then they moonlight for e-crime.
And let's go to China then. This is, you know, if we were sitting here 10 years ago, we'd be talking Kumbaya, China, we're all going to win together. The decoupling has happened essentially. Do you have operations there? Do you work with Chinese companies?
We don't have operations there. We don't have any revenue. We don't sell into China.
Never have. From the start of the company.
And what's your take on the decoupling? Let's go a little geopolitical here. They are our adversary obviously competitor. You know their technical ability is pretty strong, right? We're seeing that with AI. Are they the future competitor and the future attack vector that you have to worry about? China.
Well, I don't know that the future because I think it's already here. It's been here for the last, you know, 20 plus years. I mean, some of it may not have been talked about in the early days, but they're very very prolific. They're very good. And, you know, it's one of those you obviously nation state is always going to be nation state versus nation state. You know, that just happens. The big thing with China is there's a commercial aspect to what they do. So if the government breaks in and steals a trade secret from some company in the US, they'll just give it to another company in China, right?
Yeah. We're certainly not breaking into China to steal their trade secrets. It's going and giving them to a US company.
Yeah. The CIA is not breaking in to figure out what BYD is doing to give it to Tesla or something. Correct. So it's a one-way relationship there. Now, there's also something very unique. They make hardware. They ship a lot of hardware, our laptops in some cases, Huawei, other countries. They're putting spyware or back doors into all that hardware. Yeah. In your estimation, I mean, over the years there's certainly been cases where people have had concerns on, you know, what was actually shipped. And supply chain can be problematic whether it's actually shipped with something or, you know, in some cases packages get diverted somewhere else. The malware gets installed and it gets reshipped out, right? So, you've seen that. So it's important to have good hygiene on whatever equipment is being brought in and you know the providence of that manufacturer and that software that goes on it and that applies really around the world.
Do you trust Signal and your iPhone to protect you as a corporate executive? Like what measures do you take?
Well, because you're a target. Yeah, we're a target. I mean we our team spends a lot of time making sure that we try to do the right things. You know so WhatsApp, Signal, what about those two? You think those are attack vectors? People can do insertions there.
Well, you know I think you have to look at the privacy versus the actual attack vector. You know if you're clicking on links that are in different you know because you got a link in Signal doesn't mean the link is secure. So there's a lot of ways to get implants on these devices. I would say an iPhone is more safer than just a regular computer because the way it runs. But, you know, if you look at the nation states and some of the creative ways they can actually get into a phone, as an example, if they want to get in bad enough, they're going to get in.
How do they get in? What's the attack vector? They get you to click on a link, they insert something. That's the typical.
Yeah. Typically what they'll do, and there's a whole black or gray market on the value of a zero day. So there are zero days in all software and if you have one say for Apple iOS that's super valuable and generally if you click on something and it's not known zero day you just by going to a browser.
Explain in plain English to the audience what a zero day is.
A zero day is a yeah sorry you know talking technical jargon. Not everybody's got the hacker bonafides.
Not everybody's a hacker guy.
But yeah, if there is a software vulnerability that hasn't been found and patched, it means that they can exploit that and then they can run code. And in some cases in the past, there was a vulnerability where they could simply send you an SMS message, you didn't even know you didn't have to click on anything, but the SMS message would process this information and then they would put an implant on the phone.
Biometrics two factor has changed everything in the industry made it much more secure.
Secure but not much more. Yeah, if you look at a lot of the attacks, they're identity-based attacks. And what you see is that even when you have a credential or if you steal a credential, or if you have a session token which means that you've already authenticated.
Right. So even if you have two factor, if you have the session token, you can replay that and get access to a system.
The crypto people learned this the hard way. They basically were using two factor over SMS. People call the phone company. Reset the SIM. And then they got you and game over.
Game over. Now your Bitcoin wallet is emptied.
Where there's money, the hackers will go figure out a way.
They figure it out. And the humans are normally the weakest link.
Yeah. You know, the way I was able to get my first exploit was I called the New York Public Library. I knew they had like a VAX system with the dialup. And I just said, 'Hey, uh, it's Joe from IT at the 34th Street branch. I need the number for the dialup.' Literally, the person gave it to me at IT over the phone.
They want to be helpful. I dialed it and then I could search the stack. You could search all the books. That was the only thing you could do.
Yeah. What year was this?
Yeah. I was at Fordham and they had Bitnet. They had the internet there and I was just starting to see these things. But it's almost universally that calling somebody on the phone, meeting somebody, compromising somebody.
And normally the day the weakness is between the keyboard and the chair.
The keyboard and the chair. What is the best practice then for you when you're advising your customers of explaining to them just how the human factors part of this is the most important one to focus on?
Well, there's certainly an educational element and you also have to look at how people get paid and the motivation. So, typically they're going to call a help desk. How does a help desk get paid and what are their metrics? Their metrics is get a call, open a ticket and close it as quick as you can.
Got the incentive is speed.
Right, particularly if it's a third party in another country. So their whole job is to open a ticket, solve a problem, close a ticket. So if you follow the money and the incentives, you can see why there's a problem. And people want to be helpful at the help desk and hence they get into the situation. So you know, first start with the education piece and then having the right controls, identity protection, things like CrowdStrike Falcon is going to be additive to making sure those sort of things don't happen.
People using their own devices. Bring your own device. This is the other major attack vector. Corporations make the mistake.
Yes. Um, that you know your own device could be like a cesspool if you let your kids use it. Or it could be pretty good. But that's the reason why now there's enterprise browsers to kind of contain what people do even if you bring your own device.
Yeah. The enterprise browser is running on a virtual machine at some headquarters. You provide that kind of situation. Um, that's sort of the old school way of doing it. So what we've done, we actually just acquired a company called Sarafic.
I saw that. Explain to me the state-of-the-art there.
Yeah. So the state-of-the-art is not switching out a browser because if you use Comet or you know Firefox or whatever, you want to use the browser you have. So it's actually running beneath the browser and will support any browser which basically looks at the interaction what the browser is doing.
And if there's sort of malicious activity in the browser or there's an identity that's compromised, you can stop it at the browser and you can wall off what you do with the browser, including what data you take out of it. So it's the front door now for how people work and how adversaries get in is through the browser. How should companies implementing AI in the enterprise, putting their data into these clouds and working with partners? Claude Code this week kind of blew people's mind. I don't know if you saw the announcement or played with it.
And I was using it earlier this week and I'm authenticating on Gmail, Notion. Oh yeah, my desktop files. And I got like halfway through this and I was like, this probably isn't a good idea.
It can be pretty scary though.
Well, let me tell you a story. Not specific to that, but we don't want to throw them under the bus, but it was a lot of authentications I did.
Exactly. Exactly. They're doing great work, but there was a customer who basically created a whole suite of AI agents to help their automation, their IT department. Right. So, they had one agent that was looking for sort of IT problem, software bugs, and it found something, you know, while the code was being before it was committed. So, the agent said, 'Hey, I found this bug. I want to fix it.' But it didn't have access to fix it. So, it went to the Slack channel that had the other 99 agents and said, 'Hey, Lord, does any other agent have access to this thing because I need it fixed?' And there's an agent that raised its hand and said, 'Oh, I have access and I can fix it.' Do you see how scary this is? These two agents are reasoning and they went right around the guardrails that were put in place. This is unintended consequences and these LLMs are essentially guessing what you want them to do. They're reasoning it. Oh, it is reasonable for me to go ask for help, right? It is reasonable for me to give help. Now, what if it pushes the wrong code? What if it makes a mistake? And then, how do you ever track that down? Who's monitoring these agents? The agent technology has unlimited upside, but my lord, you're going to be in business for a long time.
Well, this is it. It's called AID, which...
Yeah. AI detection and response.
So, there's a concept that's been around that we helped pioneer, which is really endpoint detection response. So on your computer, most companies have it, you can monitor and see what's happening and prevent bad things from happening. We have to do that now with an agent. And this is why it's a huge opportunity for us because on average, each employee is going to have about 90 agents they control. So we're going to have protection and visibility across all of those agents, whether it's from a third party or whether it's a homegrown agent. And that, you know, is a massive TAM opportunity for us.
Do you trust the LLM operators with CrowdStrike's data or do you want to stand up your own large language models?
We well we have our small language models and we do work with just about all of the large language models. But we've built our own guardrails around what gets shared, how it gets shared, and how we process any of the data back. The key...
Are you concerned about it? I can see you're kind of like thinking like hm, yeah, like maybe there are some concerns here.
No, no, there is. But we tried to engineer for that.
Right. Because you want to leverage and this is the thing, you know, it's tokenomics. There's an economics to how you use the tokens as well. Plus, you want to get the right data. So, we've built small language models for things that we control with our data. And then for other bigger, you know, frontier type models, I mean, you're not going to replicate billions in R&D. You want to leverage what's out there and take the best of it.
Yeah. You're also a race car driver.
And hey, you've done well in your life. You got to buy a piece of an F1 team. I've been going to F1 races. I went to three this year. Tell me everything about it.
Where do you want me to start?
How did you get the bug? When did you get the bug? Because this has been an American thing the last couple years. I can't believe the racket these F1 you well you guys are doing. Somehow you made this the most important thing in America ever.
I knew to survive was that is a big driver and then it's a luxury experience. So now I've gone to what did I do? Miami, Austin, and Vegas. I've been to three in the last year.
I hadn't even heard about it 5 years ago.
It's fantastic. Well, I mean, little backdrop. I grew up I had zero money as a kid. I always liked...
Oh, really? I grew up in Brooklyn. Yeah.
You know, we used to play Fordham.
Okay. So, I always liked cars and fast things and never had the opportunity to do that.
First car was a Toyota Celica.
Toyota Celica. Yeah. 73 Mustang, Grande, 351 Cleveland. Okay. Okay.
351 Cleveland engine and slightly more horsepower than yours.
I know. Mine was just I had a commute in the thing.
Yeah. Okay. It was functional.
It worked. It was functional. But so I never had the opportunity to do it. And then I got into I mean I always followed sports racing. Mostly it was Indy. And then you know got into racing later. We talk about that. But I've been in Formula 1 since 2018 because CrowdStrike became well Mercedes became a customer of CrowdStrike.
And then I met Toto Wolff who runs a team and you said, 'Hey, why don't you think about working with us?'
You got in at the right time. Good. That was a good trade. It was a good trade.
So, we got in then and then over the years I built a great relationship and then just last fall I put a deal together and own 5% of the Mercedes team now.
And now you ride yourself. You're in the bronze league, I hear.
Yeah. So, I race in fact, I got to go back for a race in Daytona.
What do you drive? What kind of car?
