No Priors: Artificial Intelligence | Technology | Startups
00:08:34 · Sarah Guo
[00:08:34] S1: You have been a leader in security since, you know, Foundstone, CrowdStrike... The founding of CrowdStrike. Would you say that they're... Like, how do you put the current period of increase in capability, like attacker capability and perhaps defender capability in context? Right? Are we in a more adversary dominant world right now? Like, how do you think about the problem overall?
00:09:01 · George Kurtz
[00:09:01] S0: Well, I think it's one of the most exciting times to be in technology. I mean, really, I've been doing this thirty plus years. And, you know, what I always like to say is that security parallels the slope of the technology curve. When I first got into security, it was a Cisco router and like maybe a website barely, right? There really weren't firewalls. There were these tool kits, TIS toolkit and these other things probably dating myself, but you know there wasn't much there. Now if you think about the complexity of where we are today with with cloud, with all of the other technologies around AI, I'm not going go through all of them, but it's a lot more complex today than it was way back then. So you need all these little companies and including big companies that are able to secure all these technologies. So when we think about where we are today, it actually is the most exciting, but it also creates the biggest opportunity because the adversaries generally find the gaps or the seams. So they'll find new technologies that maybe that they don't have good protection, that they're taking advantage of things like NPM packages, right? In the old world, as you very well know, no one in corporate America would just download anything and run it on their computer. It had to be vetted and scanned and all these sort of things. And now we unleash an agent. We give it an identity, which is essentially our own or worse, and then we unleash it and we say, Do whatever you want to do. And by the way, download whatever package you want to download. And obviously these are the kind of things that cause a tremendous amount of exposure. So we're sort of getting to grips
00:10:32 · Sarah Guo, George Kurtz, Elad Gil
[00:10:32] S1: And nobody's reading any of the production code as it gets. [00:10:36] S0: No, they're not. I mean, you know, you see some of these things in a while and you're scratching your head. You shouldn't be surprised. I mean, it was like, let's give it a task and we'll come back in a few days and we'll hope it all worked. Nobody's watching it. Nobody understands what it's doing, and it's problematic. So you know we can. We can talk about sort of the explainability of what these agents are doing, which is going to be really important, but right now it's like a black box and nobody knows what's going on. They hope you know something happens and pops out on the on the other side of it, but it's a great time to be in technology. AI is fundamentally going to change the world and the frontier labs are doing amazing things, but ultimately it's going to come down to security enabling these sort of technologies. For the first time, security is the gas pedal, not the brake. And that's what we're seeing. [00:11:24] S2: What is the moment that you realized this inflection was happening, at least in the security world? And I think for a lot of people, the hugging face OpenAI issue was a big moment where people realized there's a big c shift in terms of capabilities and technology and what could happen. And to your point, are things being monitored, etcetera? Was was that a seminal moment for you, what are some [00:11:43] S0: of the seminal moments? Let me ask you a question first. So what was your big takeaway from Hugging Face?
00:11:43 · George Kurtz, Sarah Guo
[00:11:43] S0: of the seminal moments? Let me ask you a question first. So what was your big takeaway from Hugging Face? [00:11:48] S1: That anybody running frontier models needs to be careful of their environment quality and have very good ongoing production monitoring. [00:11:57] S0: Yeah, that is the takeaway. But I think what a lot of people would say as well, this was the first agentic end to end attack. All true, and you kind of look at you know what these things were doing, and you know people talk about them, you know hiding themselves and you know trying to basically survive to the next generation and those sort of things. All true. My biggest takeaway was that the defenders didn't have the same capabilities as the offense. Okay, now you would look at this and say the airbags were off because this was an experiment where the agents, you know, the guardrails were down. The airbags were off, and ultimately they found a way out of the sandbox. But the reality was Hugging Face, you know, when the first thing they did was they went to the frontier models and they said, like is this thing real? What's going on? And they got refused, right because the airbags were on over there. So what do they have to do? They actually had to go to open weight models to kind of sift through all this data and figure it out. So a big piece for me is to basically make sure that defenders have the same level of frontier defensive AI as what the offense is using. And that was a big takeaway for me and part of the things that CrowdStrike is working.
00:13:10 · Elad Gil, George Kurtz
[00:13:10] S2: And is that hard in the context that you mentioned of the guardrails being so strong on the state of the art labs models that to some extent you could argue there aren't those same defensive capabilities available to every enterprise? In other That's words, the fact that they had to go to open weight models suggests that you can't actually do it currently with the labs. And so the question then, how do you adopt that? [00:13:31] S0: Yeah, yeah. I mean, like you try to make a PowerPoint with a frontier model around cybersecurity and it refuses you. So, you know, this is part of why we created Safemind and our red and blue models just for that reason, because we wanted to give frontier caliber defenses to our customers that allow them to operate without sort of all these sort of refusals. I think that's going be an important element going forward for many companies.
00:13:31 · George Kurtz, Elad Gil
[00:13:31] S0: Yeah, yeah. I mean, like you try to make a PowerPoint with a frontier model around cybersecurity and it refuses you. So, you know, this is part of why we created Safemind and our red and blue models just for that reason, because we wanted to give frontier caliber defenses to our customers that allow them to operate without sort of all these sort of refusals. I think that's going be an important element going forward for many companies. [00:13:59] S2: I think, I guess, back to the offense versus defense side, it seems like with every technology shift, offensive capabilities are always a little bit ahead when you have a big change in how things are done for a few reasons. I mean, one of them is just with offense, it can be very directive. I need to go do x, while with defense, it's a broad surface area you need to cover. So to some extent, you need to cover more stuff. And then secondly, it's just technology adoption and diffusion. Right? You are one attack... You are a set of attackers. You know what you're going after versus the diffuse ecosystem of enterprises that you need to defend against. How does CrowdStrike think about ensuring that those defensive capabilities are very broadly available?