Good afternoon everyone and welcome back to the Cube's live coverage of Falcon 2026 here in Las Vegas. I'm your host Rebecca Knight alongside Dave Volante, my co-host and co-founder of the Cube. I would like to welcome back to the show Mike Sentonas, who is of course the president of CrowdStrike. It is always so much fun to have you on the show.
Yeah. So, I want to ask you a question before we get into all of this: what this moment is like for you in your career in terms of every board pushing AI on its companies and, "Adopt it faster. We got to go faster." And yet, we're also at this very scary time where we're seeing AI agents act in ways that are having unintended consequences, and it's pretty scary.
Yeah, I'm curious to hear in terms of your personal career and the conversations that you're having with key leaders and decision makers, what goes through your mind right now and how you see this moment in time?
Yeah, look, it's an interesting question. I've never seen anything like it, and I've never seen anything move so fast. If we go back to April 7 when we all heard about Mythos, I've never seen my phone go crazy like that with CIOs, CEOs, board members, people calling at all hours basically saying, "What do we do?" We had entire boards come into our briefing center in California because they wanted to understand what to do. I had lunch with the CEO of a really large bank. He wanted to know, was his security team doing enough and was he supporting the CISO enough because they're just so worried about what's going on. So, that's all happening on one side, and on the other side, kind of given my technical background, I think it's fascinating. It's so exciting. There's so much possibility. There's so much we can do. Unfortunately, the benefit we get is also the benefit the attacker gets, and that's kind of that challenge that we have right now.
Well, I mean, you know, we have these moments in security we've been talking about. We go back to Stuxnet or the SolarWinds hack, and now they're coming at this fast and furious pace. You know, you're going to have another one this fall. And then as a technical person, when you dig into the anatomy—I'm really glad George went into that a little bit because I think many people in this audience probably understood it, but there are a lot of people that really sort of glossed over that, and the media covered it, but not as extensively as I thought they would have. So as a technical person, I mean, what was your reaction? You guys were kind of, I guess, built for this moment. Give us your thoughts on that.
Look, I think there'd been a couple of things leading up to it that we kind of stopped and looked at and thought, "Okay, we're now getting to that point where this is happening." We've been talking about AI and the threat and the challenge of AI for 20 years really, and we're kind of the here and now, but we really needed to see the start of an attack, the start of some real-world problems. Now, the example that George talked about today with the Hugging Face incident, it was a simulation. You know, it's not as if we saw an organization get taken offline and massive, irreparable damage, but it showed us the world that we're going towards. It showed us the capabilities. It showed us that agents have the ability to work together and to cooperate for what they believe is the collective, the greater good. And it's just changing the mindset really differently. We haven't really seen it yet. We talk about Mythos, we talk about the agents. We're seeing the possibilities, and everyone's now sort of scrambling as quickly as they can to make sure that they're not the first one on the cover of the Wall Street Journal because their organization went offline for 3 weeks because of a rogue agent. But it's kind of the road. We know the road. We've seen it before. We know how the movie goes. And we're down at another chapter here of the evolution of cybersecurity threats.
Well, it unquestionably had an impact on your quarter. I mean, I can't say exactly how much in the credit pie, but there's little question that it was—I mean, you guys talked about it on your earnings call. How aware is the—I mean, obviously, your phone's ringing off the hook, the customer base is aware, but I think I heard George on Cramer say there's still a lot of companies out there that really don't understand the significance here. So, that sounds like we're still in the early days. So this thing has legs.
We are in the early days, absolutely. And keep in mind, when we talk about the Mythos moment, Mythos was announced April 7. 98% of the planet, whatever the number is—it can't be too far off—have still not seen it. People are using Opus models, they're using other models, but the Mythos moment that created the—the Mythos tool that created the moment is not generally available to everybody. People are starting to wrap their head around it, and if you look at what happened from then till now, people are hearing from other experts, other cybersecurity companies, and they're trying to prepare. So do I think this has got legs? This has got a lot of long legs because people are going to get access to new models. People are now starting to use open-weight models—open-weight models that they can download themselves that don't have guardrails, the same models that the adversaries can use, that they can train on, that they can attack with. So, this is something that we're going to be talking about for years.
How is it that you could go from endpoint to cloud to identity and now AI with essentially the same agent—sorry, sensor, we don't use the "A" term—and essentially the same platform?
We've thought a lot about this, and if you go to the early days of CrowdStrike, we thought about the challenges that had gone on for the 20 years before us. We thought about how to create this technology that would be very easy, lightweight to deploy, but importantly, this technology that would be perfect for tomorrow. That's always been the way that we thought about it. We're doing that right now, Dave. If you look at what we're doing now, Alex Ionescu was the original designer of the sensor. He's back at CrowdStrike. He's working on the next version of this technology. He's not working on today. And we keep doing that in the organization to keep building and making sure that we stay ahead of what the adversary is doing.
So when an AI agent goes off the rails, is it because it is being manipulated by an attacker, or is it because it is just relentlessly pursuing a goal in ways that weren't anticipated?
This is a fascinating question.
Yes, and it's kind of a crazy-making distinction, but—
Agents and AI inherently are goal-seeking. So when you ask the agent to do something, it's going to get you the result, and it's going to do everything it possibly can to get you the result, which is a nice thing. But in the example that we talk about with Hugging Face as an example, the agents effectively swarmed. They worked together. They chained a whole number of threats together to create a zero-day—so something no one had seen before. They created another one. So they created one to break out of the sandbox. They created another attack to break into the sandbox because they are goal-seeking. They're going to get the outcome that you asked them for. And that's on one hand really powerful for us that are using them to do something meaningful for our business, for our personal life, but for an attacker, it's terrifying on the other hand as well. And that's what we're going to have to work on solving.
Tell me about Guardian. You guys announced Guardian today. I got the pre-brief on it. Correct me if I'm wrong, but did this come out of the Pangea acquisition? Is that correct, or—
A little bit of that. So, we acquired Pangea. Pangea brought us AIDR, so this AI detection and response, prompt security. We really had a vision of AIDR not so much as a product, but we thought of AIDR as a category that includes AI gateway, it includes the prompt security, model security, the harnesses, everything we've talked about today. So, Guardian really was the vision that we had. Pangea plays part of that, but this is something that we built internally. George and I have worked really closely on that with the team. Couldn't be more proud of seeing that go live and GA today for all of our customers. It's available right now.
And then if I'm a Flex customer, how does that work?
