Defending at Machine Speed: Lisa Lu on AI Security

Summary

In this episode, Evan Cholfin interviews Lisa Lu from Stellar Cyber to explore how AI is transforming cybersecurity, the company's innovative approach to market trends, and the importance of human-AI collaboration in security operations.

Takeaways

AI-driven cybersecurity trends
Stellar Cyber's market positioning and strategy
Challenges faced by security teams today

Soundbites

"Clear messaging about AI is crucial"
"Attackers are using AI to get smarter"
"AI solutions must be battle tested"

Chapters

00:00
Introduction to AI Speed and Guest Background
01:15
The Importance of Cybersecurity in Today's World
03:24
Stellar Cyber's Unique Approach to AI and Security
05:00
Challenges Faced by Security Teams
07:06
Marketing Strategies in a Crowded Cybersecurity Market
08:47
Addressing Buyer Skepticism in AI Solutions
11:15
The Role of Human Analysts in AI-Driven Security
13:26
Defining Success in Cybersecurity Marketing
14:50
Trends vs. Hype in AI and Cybersecurity
16:52
Future Outlook for Cybersecurity and AI

Video

https://youtu.be/s6JrKL2T3qk

Transcript

unknown 0:00

Okay, okay.

SPEAKER_01 0:20

Welcome to AI Speed, the show where AI-powered companies talk about what actually works in the market right now. Business doesn't move at internet speed anymore. It moves at AI speed, and the people who figure out how to turn models into money will own the next decade. I'm Evan J. Cholfin, founder of Luxhammer, and growth partner to high-performing brands. Today, I'm thrilled to be joined by Lisa Lu, Corporate Marketing and Communications Manager at Stellar Cyber. Lisa is at the forefront of shaping how complex AI-driven cybersecurity platforms are communicated to the market, cutting through the noise in a space where everyone is claiming AI, but very few are clearly articulating real outcomes. Her work focuses on translating highly technical capabilities into messaging that resonates with security leaders, partners, and the broader ecosystem. She's also been closely involved in thought leadership and campaigns that bridge education and engagement, bringing clarity to topics like AI-driven SecOps and the evolving role of the human analyst in an increasingly automated world. Lisa, thank you for being here.

SPEAKER_00 1:32

Thank you so much for having me. Really excited for our conversation.

SPEAKER_01 1:36

Absolutely. So with all that, and you know, given your career, what originally drew you into the space, um, how do you end up focusing on uh this particular field?

SPEAKER_00 1:49

I think cybersecurity is just one of those things that everybody has to think about constantly, no matter where you started, no matter what industry you're in. If there's anything we've really learned from the last few years, it's that every company, every organization, every individual is really at risk of cyber attacks. We've seen that nobody is safe. And a lot of the tools and the philosophies that we had approaching cybersecurity in the past no longer suffice. Because as much as we're developing AI tools, as much as we create new solutions, so too are attackers also using those AI tools and they're getting smarter. They're finding new ways in. Old tools just no longer cover it. So I think because it's such a dynamic, evolving space, it's just really interesting to see how these things change.

SPEAKER_01 2:40

Absolutely. And from your marketing position, what makes Stellar Cyber's approach unique, especially in higher positioning OpenXDR in a very crowded market?

SPEAKER_00 2:50

I think the most interesting thing for me being at this company and seeing how fast trends develop is how much we've been able to predict the way the market's going, not only in the fact that everybody is saying AI. For example, at RSAC this year, there's not a single piece of collateral you'll see from any company that doesn't have AI written all over it. But we've been saying human augmented autonomous sock, and we've been integrating artificial intelligence into our story, into our product since the very beginning. So when people ask, you know, when did you start using AI, the answer is 10 years ago when we were founded, you know, and it's just it's quite amazing and rare, I think, to see your vision being validated and realized in real time by everybody around you. And it's it's you know, hard not to take a victory lap when you see all this messaging and you're ahead of the curve and you're addressing it from a new perspective and you're taking a different approach for sure.

SPEAKER_01 3:45

Yeah, absolutely. Now, I I know this may seem like an obvious question, but from what you're seeing in the market, what's the core problem security teams are dealing with today?

SPEAKER_00 3:54

The problem of a security team, I think, comes from the very nature of what security work is, which is that it's not validated by the amount of work you do, right? For other jobs, it's how many hours you put in, the things that you achieve. But for a security analyst, none of that matters until you have a breach, right? The sum of your career is pretty much the sum of your failures, the things that you fail to stop, the things that make the headlines when business grinds to a halt and there's a huge leak. So I think that entire structure of an analyst's job kind of encompasses also its difficulties because you have to be completely thorough. You have to, you know, chase every marginal addition of accuracy and visibility that you can in order to prevent the big catastrophe, right? Because that's all that anyone remembers at the end of the day. When that's when you're kind of on this defensive footing, it's easy to become overwhelmed. It's hard to tell what actually matters. Every day, security analysts face thousands of alerts, and it's very difficult to tell what actually demands their attention, what is worth the effort, and what attacks are actually going to lead to severe consequences. So the sheer volume of data, I think, is still the most pressing issue that security analysts face. And addressing those specific pain points is the goal of our product, and I think should be the goal of automated security.

SPEAKER_01 5:21

Absolutely. So, in your role, where are you personally spending most of your time right now?

SPEAKER_00 5:25

I'm spending a lot of my time doing outward-facing messaging about how our approach to AI has been more battle-tested than a lot of the other AI noise that you'll see. For example, at RSAC, where I was hands-on in the booth talking to customers, talking to prospects, somebody approached me and asked me, you know, when I'm offered an AI solution, does that mean I'm getting a tool that uses AI, or I'm getting a tool that will help me fight attackers using AI? And I just think that's a very key difference that you should be able to tell from messaging. It's very telling of where the industry is headed and also what everyone's saying that you can't immediately tell from a pamphlet or from the term AI SOC, right? So clearing up this message is where a lot of my efforts are focused right now. Telling the story of our human augmented autonomous SOC is a large part of my job.

SPEAKER_01 6:24

Yeah, definitely. And who do you think, who do you serve best? What's the ideal customer profile for Stellar Cyber?

SPEAKER_00 6:32

We have a lot of ideal customer profiles. I think because our approach is not limited to a very large, robust company with unlimited resources and the ability to hire as many analysts as they need. We're able to serve a lot of people with a lot of different needs across the market. We partner with a lot of top-performing managed security service providers or MSSPs and also mid-market enterprises. And we basically give them all the advantage of that full security team without having to spend the time or the money building that out. So because we have this integrated platform approach where we take all existing tools, we give you all the visibility you might be lacking or that you might want in addition to your existing tools, we're able to allow a smaller team to operate with the same efficiency and accuracy as a much larger one. Absolutely.

SPEAKER_01 7:25

So, what would you say is your biggest challenge this year in breaking through the noise?

SPEAKER_00 7:31

I think cybersecurity is one of those interesting industries where the playbook is constantly changing. And like I said earlier, you know, as much as we are developing new tools, attackers are too. And every day we're seeing new techniques, new things that need to be protected against. We've seen a rise in identity-based threats and these further capabilities to defend against threats that we can't even predict yet. That's a constant problem in cybersecurity, and that's also what makes it really fun, right? Like the goalposts are constantly moving. So trying to stay one step ahead instead of just keeping up, I think that's that's always the challenge.

SPEAKER_01 8:12

Yeah, definitely. And so how does that affect you from a marketing perspective?

SPEAKER_00 8:17

It means that I get to tell a different story about our about our adaptability all the time. And it's very refreshing to be able to draw upon a lot of existing messaging that we've had. For example, our messaging about keeping the human in the loop and empowering the human has just been more relevant than ever in a lot of different contexts. Every time we see the failure of fully agentic AI or unsupervised AI in any industry, not just cybersecurity, it further reinforces our messaging that the key is not to eliminate or automate the human out of the process, but rather to step up the human's game so that they're able to operate with more efficiency and have the symbiotic relationship with their tools instead of being overwhelmed by them.

SPEAKER_01 9:06

So, where do you feel the most friction right now in terms of turning interest into actual pipeline?

SPEAKER_00 9:12

I think we got, you know, a lot, we have a lot of interest. I think the last show just proved that more people than ever are aware of the gaps in their current security tools. More people are very eager for a solution that integrates all of these tools instead of just adding another one to their portfolio. Because at the end of the day, the more tools you have, the more people you need to learn those tools, to manage them, to oversee them. And a lot of people are kind of coming to the realization that that's just not a very sustainable model for a lot of organizations. So they're looking for a solution that will consolidate their current landscape and also allow them to see, you know, the whole landscape instead of just the trees.

SPEAKER_01 9:59

Yeah. Is there anything that creates skepticism for buyers when they hear AI and cybersecurity?

SPEAKER_00 10:05

A lot. I think IT is one of those industries where you have to really prove your solution because it's pretty naturally skeptical, right? As I think a lot of people are. I think technology should be battle tested before it's readily adopted. And I think a lot of the AI solutions that you can be sold these days are not battle tested. I think a lot of this chatter is very recent. And while it's exciting, it might not be enough to stake the entire future of your organization's security, right? So I think learning exactly what kind of AI is used for what purpose is an important key step. For example, AI is not a one-size-fits-all tool, right? For tool or for data normalization and correlation, LLMs are not the most efficient way to go about them. When you're asking for an AI-powered assistant, when you're looking through your security cases, sure, an LLM is really efficient, but you need the specific types of machine learning for different levels of your security operation. So that's our logic. We call it multi-layer AI. So at each step, we find the right fit instead of just throwing AI at the wall and seeing what sticks. And I think that's what differentiates our approach because we've been doing this for so many years, and because this is absolutely not new to our logic, that it's been very seamless to adopt more AI capabilities as they become available.

SPEAKER_01 11:33

And how do you communicate that to the customer so that they uh trust that that particular way or method is going to work?

SPEAKER_00 11:42

I think customers really see that for themselves too, when they're trying to adopt, you know, just any kind of AI tool, not just in their security portfolio, but in any part of their business, when they realize that sometimes, you know, the AI tool they've been given is not the most efficient way to go about things, and they start to realize there are old tools or there are other ways of doing things that might fit better. And I think those pain points really reinforce the necessity of a solution that seeks to make them more efficient instead of just selling them the most fancy or new thing on the market. Right. I think as soon as you start to experience those inefficiencies when you don't have a solution that fits your problem, that's when you realize how natural this logic comes.

SPEAKER_01 12:31

So a lot of the customers have tried other, maybe have tried other things, other solutions, and they haven't quite worked the way that it's it's been described anyway. And they come to you and you have the solution that works.

SPEAKER_00 12:48

Yeah, and I I also think that this alert fatigue is something that all analysts face. And any solution, no matter how top of the line it claims to be, that doesn't really alleviate that fatigue, ultimately at the end of the day, will not solve your problem. Our goal is to not sacrifice any of the depth of knowledge you get from a single tool, but also give you the levity to see the entire environment. And it's very hard to find that balance, and analysts feel it, right? There's no metric. It's just the way that they approach their jobs and this constant problem that security has faced from the very beginning. So yeah, I think it's because we're turning the process on its head and because we come at it with a completely different logic of how we deploy AI at every different step of our platform.

SPEAKER_01 13:40

From your position, what does success look like over the next year? What are you trying to drive from a marketing standpoint?

SPEAKER_00 13:48

Well, we would like to be the number one vendor for the top 250 MSSPs globally. We would love to obviously continue building our partnerships. We believe that cybersecurity is a team sport and that the only way to achieve more innovation is just through collaboration. We're so excited to see what all of our partners have brought to the table, what we're able to do together. And I think just growing our ecosystem and continuing to deliver the best solutions to the people who depend on them. That's what success looks like.

SPEAKER_01 14:19

Yeah, absolutely. What's what's actually working right now on acquiring customers and cybersecurity?

SPEAKER_00 14:25

I think a lot of our customers see the value of our approach from the very beginning because we approached the question from really the point of view of the analyst instead of you know just hopping on whatever train is passing by. After all, we didn't just get on the AI train, we've been conducting it from the very beginning. So I think that perspective shift and the fact that we're driving the symbiotic relationship between a tool and an analyst instead of just building a tool, that's exactly what draws partners and customers to our product.

SPEAKER_01 15:03

Yeah, absolutely. What trends in AI and cybersecurity feel real to you right now versus just hype?

SPEAKER_00 15:09

Well, everyone's saying agentic AI, again, not really. Everybody knows what that means. I think there is a desire to move to fully autonomous operations where AI agents can do everything unsupervised. But the thing to also keep in mind is that when a human makes a mistake, it's a mistake. When an AI agent makes a mistake, it can it can snowball into something absolutely catastrophic because they're so much more efficient, because they operate, you know, a thousand times faster than a human can, which also means that their mistakes can cost you way more than a human analyst's mistake. Which means that I don't think that we are able to go fully hands-free yet. You know, as we saw from a power outage in San Francisco, a lot of the Waymo's short-circuited and we're blocking the streets. What do you do then? You need a human driver to climb behind the wheel and get all the cars out of the street so that everything can go where it needs to be, right? And so that's that's the logic with our human augmented autonomous talk is that we can't take the human out of the loop. The human needs to be strategically placed so that they're not wasting their energy doing the grunt work that, you know, exhausts them and is also not the most valuable use of their time, but that they're able to learn from the AI models and become smarter and better at their jobs themselves while making sure that everything is running as it should. So, and this is also essential in building trust in the technologies, right? No one is going to adopt a solution that they haven't seen for sure works. So this is kind of, you know, the way things are moving instead of the way I think a lot of people are projecting they'll move.

SPEAKER_01 16:51

Yeah, absolutely. Well, I used to ask where do you see the market going over the next 12 to 18 months? But these days, I think the next three months is probably more appropriate to ask. So, where do you see the next three months going?

SPEAKER_00 17:06

Yeah, I see attacks moving in a much faster direction than we thought possible. I see, you know, attackers using Claude as well to upscale and to make themselves more efficient. I see in cybersecurity, really, you can't stay in one place for longer than six to nine months. It moves extremely quickly. And as soon as you think you found a solution to one kind of threat, a new one pops up and it's going to be faster and smarter. So the way AI agents are moving through the security market is also reflective of the way AI agents are moving through the other side. So I think this constant game of keep staying one step ahead, trying to catch up, trying to predict where attackers are going to go next, that's that's the way things are gonna keep going for the next few months.

SPEAKER_01 17:58

Yeah, definitely. So if we were to have this conversation again in a year, what would need to happen for you to feel like it's a big win for you guys?

SPEAKER_00 18:08

We've got a couple more shows coming. We're excited for Black Hat in August. That's in a few months. We'll be really excited to see everyone there. And we would just like to keep serving our customers and to keep growing our community and to keep, you know, having our vision validated by the various awards we've won. We were named most innovative in the AI cybersecurity awards by Global InfoSec at RSAC 2026. So more validations of our vision, you know, continuing to evolve our product, offer more capabilities, and streamline operations. That would be that would be what success looks like, I think.

SPEAKER_01 18:46

Yeah, no, I think that'd be huge. Well, that's it for today's episode of AI Speed. A huge thank you to Lisa Liu for sharing her invaluable insights into how Stellar Cyber is cutting through the noise and helping security teams focus on real outcomes with AI-driven operations. If you're building or leading an AI native company or a service business that uses AI under the hood, and you care about revenue, adoption, and market share, make sure to subscribe to AI Speed. Learn how the best AI operators ship faster, sell smarter, and stay ahead. Thanks for listening. Until next time, keep building, keep selling, and keep moving at AI speed.