In The Arena by TechArena - How Palo Alto Networks is Fighting AI Threats with AI

Episode Date: March 6, 2025

In this episode of In the Arena, Palo Alto Networks’ Dharminder Debisarun explores the challenges of securing smart industries, preventing attacks, and staying ahead in an evolving threat landscape.

Welcome to the Tech Arena featuring authentic discussions between tech's leading innovators and our host, Allison Klein. Now let's step into the arena. Welcome into the arena. My name is Alison Klein. We're coming to you from MWC, and I am so grateful to be with Dharminder Debusaroon, Smart Industries Cybersecurity Executive at Palo Alto Networks. Welcome to the program. Thank you, Alison.

So Dharminder, you've never been on the show before. Why don't we just start with an introduction of your history in the industry and a little bit about Palo Alto Networks. So I'm now nine years with Palo Alto Networks. Palo Alto Networks is a cybersecurity company. And what we do or what I do in my day-to-day business is share our insights how we can secure smart industries. Everything that's basically OT, IoT related within Palo Alto Networks, we call it smart industries. Everything that's basically OT, IoT related within PowerAlternate, which we call it smart industries. Now, I know that you have a very deep history in IoT

and Edge is the topic of 2025 for a lot of reasons. Edge is growing rapidly. What are some of the prevailing use cases that you're seeing when you're talking to customers that's driving that adoption curve? One of the prevailing use cases that you're seeing when you're talking to customers that's driving that adoption curve? One of the biggest use cases, basically in every industry that we see, has to do with predictive maintenance.

Predictive maintenance is key because when you are working on production lines, basically you want to continuously make money because everything is about making money in the industry. So downtime maintenance is not acceptable because it needs to work 24 by 7. So what you want to do as owner of these production lines, as the plan manager, know exactly how can I stretch this. And when you are using IoT with all these sensors and getting all these data, then you basically know, okay, this is the moment that this production lines or part of this production lines would need maintenance.

And then you can really plan that and manage that. So that's one of the biggest use cases that we see. But when we look at the automotive industry, IoT is a huge one, right? Take the radar's LIDAR systems. All these autonomous vehicles that are basically on the market are using a lot of AI. And IoT is a huge part of this. Now, when you look at AI proliferation, a lot of the edge use cases for AI have been around traditional AI ML models.

Obviously, 2024 was occupied with conversations about gen AI. What is actually happening in the edge today? In the edge today, what we see is cybersecurity is getting really interested in this complete AI landscape, especially on the Edge is how are we basically going to secure this environment? How does cybersecurity play a role? And there where we as Palo Alto Networks, one of the first pioneers with AI

in cybersecurity come in place. What we do is fight AI with AI. So we have to across portfolio, we have embedded cybersecurity AI. So what does that really mean? Where we are heading as industry, especially focusing from the edge to cybersecurity is more about these AI angels. They are design, develop, machine learning, large models. The A need to be protected against infiltration. So you don't can misuse the parameters and then an outcome will be different.

Right. And that's what we train our AI models to understand how does the AI work and what can we do against this protecting this. Looking at digital twins where you have this digital twin of your production line and what if parameters are being changed. You cannot do that in the production environment. So you need the digital twin to tweak and tune your production line. So if that parameters are changing, you can perform much better.

Basically you didn't have to do it in a lab. But that detail needs to be also protected. So cybersecurity across the whole industry for the calls is massive at this point. Now you spoke yesterday at Mobile World Congress on operational security. What was your message to the audience and what do you think is new in terms of the feedback that you're getting at MWC? So what's new? Because that's very interesting always to decide, hey, what haven't we seen before? What we see is that the IT-OT convergence still is difficult to manage. And if we look at our latest research, over 72% of the attacks are coming from IT to OT.

So we still need to think about how we can implement some kind of zero trust, right? I don't want to say fully zero trust, but think about this type of strategy, frameworks that can help you secure your environment. Because 100% security doesn't exist, but what we can do is make it so hard to get in. And once you get in, try to stop lateral movement. Because that's how it works, right? OT is complete lateral movement. So if you can stop lateral movement and contain that, I think that you have done your work.

Because if an attacker wants to get you, they have more facilities available that you can basically use to create yourself. And that's what we always need to remember. Hey, if I'm our big player, you need to notice, hey, it can happen. And what I see is missing, especially on C level is what's the playbook if you are compromised? And nine of the 10 times I call my lawyer, you'd call first looking at your own team, hey, what information I'm getting, how are we going to communicate? There is so poor information within the C level and board levels available

to go to that once it happens. You know, it's interesting that you say that. I was going to ask you, what are the steps, because some of this is technology, but some of it's human. So what are the steps that business leaders need to take to prevent these cyber attacks? So there are, of course course regulations, right? In Europe, we have this too, that really now say, hey, you as a CEO are now responsible for any cyber attacks happening.

But I think one of the best thing to do, especially on board level, is realize it can happen to us. So what needs to be in place, right? How can we defend ourselves? Because it's just a forensic investigation that will happen and somebody will say, hey, did you manage the deadlines, 24 hours, et cetera, et cetera. Besides that, of course, you have insurance and those kinds of things, but I don't want to go to that direction.

What I want to say to business leaders is, how does the process look internal? When will you communicate what? And do a drill, do an exercise, right? We are compromised. Bring the playbook for me. And then you can basically go step by step. You don't have to invent the wheel. We have enough consultancy companies that can help you with that.

We have enough consultants in companies that can help you with that. But every wording, every sentence that you bring out can have consequences. And you need to be very careful in our industry about bringing something upfront because no company likes to be on the headlines of being attacked, right? So a proper tabletop exercise would help them. So we do a lot of tabletop exercise. So we create a communication. No answer is wrong. We give you only points.

The points are less. And when you do tabletop exercise, no one wants to do back, right? So with that angle, we created the game. Let's give points and based on the points that you are getting on the first question, the question will get easier or difficult. So that end results for everybody, what we say, right? So the mathematics behind this. And then you see different approaches in the same room.

And let's say almost everybody is part of a management team or a board team. And they have only different angles that they think is important for them. You know, in a previous life, I took part in one of those tabletop exercises around a ransomware scenario. And it was fascinating what we learned collectively from just going through that mental process. And every time it comes to communication, we see so bad communication between levels below

to the top that basically the top, okay, what do we need to say? And then somebody technical is dictating something and somebody is saying, it doesn't make sense if you don't know what you're talking about. Yeah, and communication in those moments is so important. Now Palo Alto Networks also made a splash in terms of what's new. You guys are known for enterprise security, but you came up with an announcement here that I thought was quite interesting.

Do you want to tell us about it? Yeah, so we did several announcements, but the biggest announcement that we have done is on the 5G security landscape. So what we have combined is 5G with SASE. So we give service providers the opportunity to monetize with security on top of it to this enterprise users. So we can do, for example,

same authentication in a SASE concept. So we have been doing this, trying to enter this market for a while, right? Because we hold a lot of patterns on the 5G security part. But what we saw is so hard for the service providers to sell it to the engines and now with this concept, we give basically an entry to all the service providers to sell their private LTE or with SASE concept much easier. Because you have constantly security everywhere.

And that's something customers are not aware of, right? There's so many discussions going about, I'm using this application, it shouldn't be secured. Yeah, but your data, it's about the data. Where's the data located? What measures are you taking to secure that data? Yeah, but I thought the cloud provider, the SaaS provider is doing that, but that's not always the case. You need to think about, and of course, a long-term relationship with NTT and the other, so we added more partners that are able to work with Palo Alto networks with our OT IoT security solutions, where we're going to go big globally. That's really awesome.

Now, one question that I have, we were talking about bad actors before. Obviously, bad actors have gotten more sophisticated with AI as well. How do you see that landscape changing and how are you keeping ahead of that? Yeah, I always say it's a cat and a mouse. A bad actor needs to be right once. We need to be right always. And that's a huge difference. Little pressure.

A little bit pressure. Can you imagine what will happen if we are compromised? Yeah. Oh, that would be lots of good. Right? With around 85,000 customers and more, it will be devastating. What we do as one of the pioneers in AI cybersecurity, because we have been doing it for so many years, we always try to be a step ahead of the game.

And how we do that is investing a lot in R&D. As a cybersecurity company, you need to always know how does the bad world look like. So we invest a lot in R&D, looking at the bad world, looking at what's the next big thing. And then, okay, what can we do towards that? Because what we see a lot of errors being made by humans. Somebody forget to close something or wide open, convey change, et cetera. So what we are trying to do is basically tackle that bar to help the industry.

Everybody's releasing co-pilot, but you can imagine that there will be a time that making policies on the firewalls or other systems can be fully automated, that you only drag and drop where needs to communicate to work. What's the source destination and rules are being created. to work, what's the source destination, and rules are being created. So taking away the human as much as possible, automation is key at that, will create, I say, a better way to secure environments, but the bad world would have to make more efforts than they already do to get in, because we need to be clever. We need to be a step ahead of the game.

And automation is the only way we can do it. If we're going to use human intervention, it will cost time. When you are under attack, you don't have time. When somebody wants to pivot in your environment, what he's going to do is he's not going to immediately go and try to attack you. He will listen on the net, see what's passing, what's relevant, what's not relevant. And in time windows and really small packages, information will be sent to the outside of your network without realizing.

And IoT is a huge stepping stone for these guys, because you can't install something on an IoT device, right? So what you need to do is use a lot of behavior analytics and anomalies, those types of deviation to know, hey, what's going on in my environment? And the biggest challenge with IoT, no standard applications, supply chain, just build hardware and software from somebody else. So many things that you need to think about. And you can't do that with humans.

You need visibility, you need AI to help you with that. You will see in the upcoming years, a lot of investment from cybersecurity companies in AI detection, AI prevention, et cetera. One final question for you. It's been fascinating to listen to what you're delivering. I know that people are going to think, hey, what are all the other announcements from Alalto Networks and I want to read about them.

Where can they go find out more information and where can they connect with you? Yes, of course they can connect me with a mobile app from the Mobile World Congress, but on our website, parlouralternetworks.com, you can fill out all the information, all the contact details from personal or you want to have local reach out. You will find everything on the Parla Auto Networks website. Thank you so much for taking some time from your very busy MWC schedule. It was really delightful to get to know you a bit. Thank you so much for taking some time from your very busy MWC schedule. It was really delightful to get to know you a bit. Thank you.

Thanks for joining the Tech Arena. Subscribe and engage at our website, thetecharena.net. All content is copyrighted by the Tech Arena. You