Latest / Expert Insights Podcast / #85 - OpenText World Roundup
Transcript
- 0:05SPEAKER_02: Hello and welcome to the Expert Insights Podcast.
- 0:11SPEAKER_02: I'm Meryn McDead, reporter at Expert Insights.
- 0:14SPEAKER_02: Over the past week, I've been in Nashville attending OpenText.
- 0:17SPEAKER_02: I've been attending meetings,
- 0:19talking with industry experts and exploring how the industry
- 0:21will evolve in 2026.
- 0:22SPEAKER_02: In today's podcast episode, I've got two interviews from the conference.
- 0:29SPEAKER_02: One from Scott Richards,
- 0:31SPEAKER_02: SVP,
- 0:32SPEAKER_02: AI and Discovery and Cybersecurity,
- 0:34SPEAKER_02: and the other from Muhi Mazou,
- 0:36SPEAKER_02: EBP of Cybersecurity Engineering at OpenText.
- 0:41SPEAKER_02: Let's dive into our first conversation with Scott,
- 0:44where he discusses the rapid expansion of agents and technology
- 0:47and the impact that
- 0:48SPEAKER_02: this is already having.
- 1:01SPEAKER_02: Hi, Scott, and thanks so much for joining me today.
- 1:04SPEAKER_02: I'd like to start off by asking you to introduce yourself,
- 1:08SPEAKER_02: your role at OpenText,
- 1:09and maybe go into a bit about how your work intersects with
- 1:13emerging AI-driven
- 1:14SPEAKER_02: engineering practices.
- 1:16SPEAKER_00: Sure.
- 1:17SPEAKER_00: So my name is Scott Richards.
- 1:18I'm the senior vice president of product and engineering for
- 1:21our cybersecurity enterprise group.
- 1:24So we also have a cybersecurity SMB and consumer team,
- 1:27but I run the enterprise team.
- 1:28We have,
- 1:29we kind of divide our enterprise cybersecurity group into five
- 1:33pillars.
- 1:34SPEAKER_00: So those five pillars are security analytics or SecOps.
- 1:38SPEAKER_00: We also have an application security team.
- 1:41SPEAKER_00: We have a data security team.
- 1:43SPEAKER_00: We have an identity and access management group.
- 1:47And then the last one we have is called Digital Forensics and
- 1:50Incident Response,
- 1:51which is kind of a unique category that's really focused on
- 1:54really hardcore,
- 1:55SPEAKER_00: deep forensics leveraged by law enforcement,
- 1:58SPEAKER_00: intelligence agencies,
- 2:00SPEAKER_00: people like that.
- 2:01SPEAKER_02: Excellent.
- 2:01And then your second part of the question was on kind of
- 2:05AI-enabled solutions.
- 2:07SPEAKER_00: Is that...
- 2:09SPEAKER_02: how your work intersects with emerging AI-driven engineering practices.
- 2:13SPEAKER_00: Yeah, yeah, it's interesting.
- 2:15I think it doesn't just intersect with it,
- 2:18it's almost a complete and total overlap with it.
- 2:20SPEAKER_00: So we are extremely focused,
- 2:23SPEAKER_00: as is the industry,
- 2:24SPEAKER_00: not just securing against the typical human users or insider threats,
- 2:30SPEAKER_00: but now we're having to worry about how do we secure agents,
- 2:33SPEAKER_00: right?
- 2:33SPEAKER_00: Mm-hmm.
- 2:35SPEAKER_00: In some reports, agents are already outnumbering humans by like 50 to 1.
- 2:41So you can imagine a CISO at an organization who's been
- 2:46stressed about identity
- 2:48SPEAKER_00: management,
- 2:48SPEAKER_00: data access,
- 2:49SPEAKER_00: et cetera,
- 2:50SPEAKER_00: with all of the humans doing crazy,
- 2:53SPEAKER_00: dumb,
- 2:53SPEAKER_00: sometimes malicious things.
- 2:55Now multiply that by 50,
- 2:57trying to figure out what our agent's going to do in this
- 2:59world.
- 2:59SPEAKER_00: So we are right there at that intersection.
- 3:02Everyone we talk to from a customer perspective is concerned
- 3:05about their data,
- 3:07SPEAKER_00: how agents are accessing that data,
- 3:09SPEAKER_00: how are we going to,
- 3:11SPEAKER_00: how identities to those agents so that we can apply policies,
- 3:15SPEAKER_00: manage access,
- 3:16SPEAKER_00: et cetera.
- 3:17SPEAKER_00: Mm-hmm.
- 3:18SPEAKER_02: I mean, it's all quite new, isn't it?
- 3:19SPEAKER_02: So, of course, people are going to have questions and concerns.
- 3:22SPEAKER_02: It's very understandable.
- 3:23SPEAKER_00: Yeah, exactly.
- 3:25You delivered a keynote yesterday that went into AI agents
- 3:28becoming the new attack
- 3:31SPEAKER_02: surface for businesses.
- 3:32For our readers,
- 3:33what were the key insights or takeaways that you covered?
- 3:36Yeah,
- 3:38so agents are proliferating and there's nothing we can do about
- 3:43it.
- 3:43We could potentially just say,
- 3:46hey,
- 3:46we're going to outlaw or ban agents from our organization.
- 3:50SPEAKER_00: As I mentioned in my keynote,
- 3:51SPEAKER_00: that would be a little bit like saying,
- 3:53SPEAKER_00: hey,
- 3:54SPEAKER_00: flying has become more risky,
- 3:56so we're just going to ground airplanes and we're never going
- 3:58to fly again.
- 3:59SPEAKER_00: Clearly, that's not the solution.
- 4:01You would be at a competitive disadvantage immediately if you
- 4:04chose to do that
- 4:05because these agents are extremely helpful to do a lot of
- 4:08different things.
- 4:08The question then becomes,
- 4:10how do we make sure that we are safe with these agents?
- 4:14And I think there's a number of steps and we feel strongly that
- 4:18we're very well
- 4:18positioned at OpenText because we have a breadth of solution
- 4:22that kind of cover the
- 4:23SPEAKER_00: entire attack surface.
- 4:24SPEAKER_00: And what's critical is that with these agents,
- 4:27SPEAKER_00: We have to really focus on preemptive security.
- 4:30Used to live in where we could just watch for known attack
- 4:33patterns and then alert
- 4:35SPEAKER_00: someone when that pattern was recognized.
- 4:37SPEAKER_00: We still need to do that.
- 4:38SPEAKER_00: That's still a valid security response.
- 4:41SPEAKER_00: However, it's not good enough anymore and it's not sufficient.
- 4:44We have to do more and we have to do more at the front end of
- 4:47the attack sequence.
- 4:48SPEAKER_00: So these attacks just don't happen.
- 4:49SPEAKER_00: So things like just to give you a few examples.
- 4:53We need to use very strong AI-powered application security
- 4:57solutions to look
- 4:59SPEAKER_00: through the code and find vulnerabilities in the code of these agents,
- 5:03either in the agents themselves or in the applications that
- 5:05these agents develop,
- 5:07to make sure that we're filling gaps where things like prompt
- 5:11injection can be
- 5:12SPEAKER_00: inserted,
- 5:12SPEAKER_00: where things like excessive agency can be exploited.
- 5:15SPEAKER_00: So first step, making sure those agents are secure.
- 5:18SPEAKER_00: We need to make sure that we are locking down our data.
- 5:22SPEAKER_00: I mean, the data is what is exposed.
- 5:23SPEAKER_00: That's what these bad actors want.
- 5:26They either want to block our access to that data or they want
- 5:29to gain access for
- 5:30SPEAKER_00: themselves and exploit it.
- 5:32SPEAKER_00: You know,
- 5:33doing things like making sure that you have a solution that
- 5:37goes through the entire
- 5:38SPEAKER_00: ecosystem,
- 5:39SPEAKER_00: identifies PII,
- 5:41SPEAKER_00: identifies PHI,
- 5:42SPEAKER_00: other sensitive information,
- 5:44SPEAKER_00: and then not only taking measures to protect it through encryption,
- 5:48SPEAKER_00: etc.,
- 5:48SPEAKER_00: but also reporting that into other security solutions.
- 5:52SPEAKER_00: So this is a team sport, right?
- 5:54SPEAKER_00: There is no one silver bullet that's going to solve this problem.
- 5:57So we have to make sure that our data security solution is
- 6:00reporting to our threat
- 6:01SPEAKER_00: detection and response solution that,
- 6:03SPEAKER_00: hey,
- 6:03SPEAKER_00: there's sensitive data in this file,
- 6:05SPEAKER_00: in this repository.
- 6:06SPEAKER_00: So that solution can pay closer attention to those.
- 6:09We do some very unique things in threat detection where we are
- 6:12monitoring behavior.
- 6:14SPEAKER_00: We've done it for a long time,
- 6:15SPEAKER_00: monitoring behavior of humans,
- 6:18SPEAKER_00: where we figure out the unique normal,
- 6:21SPEAKER_00: right?
- 6:21SPEAKER_00: So the unique normal for me might be I work from Utah.
- 6:24SPEAKER_00: I usually work between 6 a.m.
- 6:25SPEAKER_00: and 7 p.m.,
- 6:27So if all of a sudden I'm doing something from Africa at two in
- 6:30the morning,
- 6:31SPEAKER_00: there's probably something that's gone wrong.
- 6:33SPEAKER_00: And we can do that same thing for agents.
- 6:35SPEAKER_00: So watching and baselining their unique timestamps for when they work,
- 6:41SPEAKER_00: what they access,
- 6:42SPEAKER_00: how often they access certain files.
- 6:44SPEAKER_00: So when that behavior becomes atypical,
- 6:46we can immediately alert and notify before damage is done in an
- 6:51attack sequence.
- 6:52And then that feeds into identity and access management where
- 6:54we can lock down
- 6:56SPEAKER_00: their access to certain folders,
- 6:57SPEAKER_00: files,
- 6:58SPEAKER_00: etc.
- 6:58SPEAKER_00: So it's this continuous loop of dynamic security that evolves.
- 7:03SPEAKER_00: And I think it's two things are really important there.
- 7:05SPEAKER_00: One, it has to evolve.
- 7:06SPEAKER_00: So it has to change because the threats are dynamic.
- 7:09SPEAKER_00: The threats are changing.
- 7:11SPEAKER_00: And then number two, it has to be AI powered.
- 7:14There is just no world anymore where humans can keep up with
- 7:18the advanced threats
- 7:20SPEAKER_00: that are being thrown at us.
- 7:21SPEAKER_00: We have to do this at machine scale.
- 7:23We've got to be able to sort through billions and billions of
- 7:26events and correlate
- 7:27those events and then identify these are the three things that
- 7:30went together,
- 7:31SPEAKER_00: represent a risk,
- 7:33SPEAKER_00: and then take appropriate response there.
- 7:35That is something I've noticed in a common theme of people
- 7:37discussing AI and how
- 7:39SPEAKER_02: you just,
- 7:39SPEAKER_02: you can't be left behind.
- 7:41SPEAKER_02: You've got to keep rolling with good times, as always.
- 7:44SPEAKER_00: 100%.
- 7:45SPEAKER_02: So when you look across the enterprise,
- 7:48what specific AI-driven risks should organizations be most
- 7:51concerned about right
- 7:52SPEAKER_02: now?
- 7:53SPEAKER_02: So we've got the list,
- 7:54SPEAKER_02: insecure code generation,
- 7:55SPEAKER_02: data exposure,
- 7:57SPEAKER_02: workflow hacking,
- 7:58SPEAKER_02: or is it something else entirely?
- 7:59SPEAKER_02: Yeah.
- 7:59SPEAKER_00: Yeah, I think those three are great.
- 8:01SPEAKER_00: I would identify a couple of others as well.
- 8:04SPEAKER_00: I think this kind of the world of agent and agent proliferation,
- 8:08SPEAKER_00: it's related to those that you just listed.
- 8:10We're in this unique new position where we have to decide how
- 8:13we're going to manage
- 8:14SPEAKER_00: these agents because it's the wild,
- 8:15SPEAKER_00: wild west right now.
- 8:16SPEAKER_00: People are building their own agents.
- 8:19SPEAKER_00: We have,
- 8:19we announced and talked about an aviator studio that's going to
- 8:22allow people to
- 8:23SPEAKER_00: create aviators,
- 8:23SPEAKER_00: which are agents.
- 8:24We're doing it in a safe and secure,
- 8:26organized,
- 8:27orchestrated way,
- 8:28which is what we need to do.
- 8:29SPEAKER_00: But when you think about agents,
- 8:30SPEAKER_00: you've got,
- 8:31SPEAKER_00: you've kind of got the digital assistant,
- 8:33SPEAKER_00: which is my identity.
- 8:35SPEAKER_00: It's doing my bidding.
- 8:37SPEAKER_00: It's doing what I want it to do.
- 8:38SPEAKER_00: Relatively easy to control because it's doing what I ask it to do.
- 8:42SPEAKER_00: You've got the digital worker that is separate, right?
- 8:44SPEAKER_00: It's a separate entity all by itself.
- 8:46SPEAKER_00: It's autonomous.
- 8:47SPEAKER_00: It's doing what it wants.
- 8:48We can assign it an identity and make sure that it stays inside
- 8:52the scope we intended.
- 8:53Where it gets really scary is when you talk about things like
- 8:56digital twin,
- 8:57right?
- 8:57SPEAKER_00: That's my identity,
- 8:59SPEAKER_00: but it's working autonomously,
- 9:02SPEAKER_00: hopefully doing what I want it to do,
- 9:04SPEAKER_00: but no guarantee it's not going to go out of scope or
- 9:08SPEAKER_00: It's not going to always perfectly interpret what I want it to do.
- 9:11We've got to,
- 9:12as an industry,
- 9:13figure out how are we going to identify these?
- 9:16SPEAKER_00: How are we going to assign identities?
- 9:18SPEAKER_00: How are we going to manage their access?
- 9:19SPEAKER_00: How are we going to monitor what they've touched?
- 9:21SPEAKER_00: And we're well positioned at OpenText to do that.
- 9:24SPEAKER_00: We've been doing identity and access management for two decades,
- 9:27SPEAKER_00: but we have to evolve that to also now monitor agent behavior,
- 9:32SPEAKER_00: which is really interesting and exciting.
- 9:34SPEAKER_02: That is very interesting.
- 9:35SPEAKER_02: We're seeing an explosion of AI security solutions right now.
- 9:40AI coding protectors,
- 9:41agent identity government tools,
- 9:43data safety layers and more.
- 9:46Is OpenText aiming to kind of create an end-to-end security
- 9:50pipeline that covers
- 9:51SPEAKER_02: all of these areas for enterprise AI?
- 9:53SPEAKER_02: What is your vision of the future for AI security?
- 9:57SPEAKER_00: Yeah, that's a great question.
- 9:58SPEAKER_00: I would say the short answer is no.
- 10:00So I think it would be very egotistical of any organization to
- 10:05think they were
- 10:05going to be able to provide a solution across every facet of
- 10:08the attack surface.
- 10:09SPEAKER_00: That would be, if somebody is telling a customer that
- 10:13SPEAKER_00: I think their antenna should go up.
- 10:15SPEAKER_00: They're probably being sold snake oil.
- 10:18SPEAKER_00: No one can do that.
- 10:19SPEAKER_00: And honestly, no one should try to do that.
- 10:21SPEAKER_00: So we are very,
- 10:23very good at several aspects in the security sort of threat
- 10:28surface or threat
- 10:29SPEAKER_00: protection.
- 10:30SPEAKER_00: We're very good at identity and access.
- 10:32We've been in the Gartner magic quadrant,
- 10:35leading quadrant for 11 years in application security.
- 10:38SPEAKER_00: We're great at that.
- 10:40We have format preserving encryption that makes us unique in
- 10:43the data security world.
- 10:45SPEAKER_00: are things that we don't do.
- 10:46SPEAKER_00: And that's where, you know, I mentioned earlier, this is a team sport.
- 10:50SPEAKER_00: And if we're going to win against the threat actors,
- 10:54SPEAKER_00: the malicious actors that are spending millions and millions of dollars,
- 10:58SPEAKER_00: billions of dollars,
- 10:59there's entire universities in places that are only teaching
- 11:03people how to do bad
- 11:04SPEAKER_00: malicious things.
- 11:05If we're going to win against that,
- 11:07we've got to work together as an ecosystem.
- 11:09SPEAKER_00: So, you know,
- 11:10we we have our strengths we partner with in areas where we
- 11:15don't have strengths and
- 11:17SPEAKER_00: we are perfectly happy to do that because the reality is there's other
- 11:20organizations that are better at us at certain things and we'll
- 11:24partner with them
- 11:25to try to create this cooperative fabric of security solutions
- 11:29that work together
- 11:30feed off of each other to provide this layered security that's
- 11:34what it's going to
- 11:34SPEAKER_00: take
- 11:35SPEAKER_02: Yeah, makes perfect sense.
- 11:36SPEAKER_02: So from an engineering leadership perspective,
- 11:39how do you balance enabling AI innovation with enforcing strict
- 11:43security controls?
- 11:44SPEAKER_00: Yeah,
- 11:45SPEAKER_00: that's what every industry,
- 11:47SPEAKER_00: every engineering organization is kind of battling with,
- 11:50SPEAKER_00: right?
- 11:51You can't be last in this game or your engineering teams will
- 11:55be frustrated and
- 11:56SPEAKER_00: leave because they want to go to more innovative,
- 11:58SPEAKER_00: more cutting edge organizations that allow the use of technology.
- 12:04SPEAKER_00: We're probably not first either.
- 12:07SPEAKER_00: We are a massive organization with thousands and thousands of developers.
- 12:11So we have to have some controls in place,
- 12:14some guidance in how that technology is used.
- 12:17SPEAKER_00: But we're certainly at the front edge.
- 12:19SPEAKER_00: We're certainly leading in that.
- 12:21So our developers are both using our own tools for AI powered
- 12:27development initiatives.
- 12:28We talked about application security that's deployed across all
- 12:31of OpenTech.
- 12:32So we use our aviators or our agents to not only look through
- 12:36code and find gaps
- 12:38SPEAKER_00: that might allow for things like we talked about,
- 12:40SPEAKER_00: whether it's prompt injection or whatever.
- 12:42SPEAKER_00: And then we use our aviators to actually
- 12:45provide automatic code suggestions so our developers can
- 12:49literally just click a
- 12:50button and entire code snippets are in the ci pipeline or
- 12:54channeled into their
- 12:54development environment and replace the vulnerable code massive
- 12:58massive time
- 12:58savings we're also leveraging tools from partners right so
- 13:02we're absolutely in that
- 13:04process we're deploying and becoming more active there every
- 13:08day but we're
- 13:08SPEAKER_00: definitely taking a measured organized approach as opposed to just
- 13:12SPEAKER_00: letting everyone go rogue and do kind of whatever they want.
- 13:15SPEAKER_00: That's not a development environment that I think we're comfortable with.
- 13:19SPEAKER_02: No, definitely got to find that balance, can't let anyone go rogue.
- 13:22SPEAKER_00: Exactly.
- 13:23SPEAKER_02: So looking ahead to 2026,
- 13:25what advice would you give to security leaders to prioritize in
- 13:29terms of AI
- 13:29SPEAKER_02: security or beyond?
- 13:31SPEAKER_00: Sure.
- 13:31SPEAKER_00: I think number one, it's about the data.
- 13:34SPEAKER_00: So my number one piece of advice would be make sure your data is
- 13:38is secure and it's ready for this proliferation of agents
- 13:42because it's coming.
- 13:43SPEAKER_00: If you don't think it is,
- 13:44SPEAKER_00: then it's probably happening behind your back,
- 13:46but there are going to be agents in your environment and they
- 13:49need to make sure
- 13:50SPEAKER_00: that every corner of their content,
- 13:53SPEAKER_00: every corner of their data repositories is secure.
- 13:56and that means scanning those environments identifying where
- 14:00they're sensitive data
- 14:01and then taking action whether that's redacting that data
- 14:04masking that data
- 14:05relocating that data or encrypting that data and and we have
- 14:08solutions that do all
- 14:09of those things so when a bad actor accesses the data they're
- 14:14either getting the
- 14:16false information that you want them to get or they just don't
- 14:18have access at all
- 14:19SPEAKER_00: so number one
- 14:21SPEAKER_00: Make sure your data is secure.
- 14:23SPEAKER_00: Number two,
- 14:24you need to make sure that all of these agents that are being
- 14:27deployed have an
- 14:28SPEAKER_00: identity and you're managing them,
- 14:31SPEAKER_00: watching them,
- 14:32SPEAKER_00: monitoring them just like you would any human.
- 14:34SPEAKER_00: So that's number two.
- 14:35And then the third thing I'd say is make sure that you are
- 14:39monitoring the behavior
- 14:41of those agents so that you know when they are going outside of
- 14:45the bounds which
- 14:45SPEAKER_00: you've set for them.
- 14:46SPEAKER_00: So I don't think we should be afraid of leveraging agents.
- 14:50I just think we need to be organized and take the precautions
- 14:53necessary to kind of
- 14:54build that runway so that we can safely deploy these things and
- 14:58do so in kind of an
- 15:00SPEAKER_00: organized manner.
- 15:01SPEAKER_02: Excellent.
- 15:03That's all my questions for today,
- 15:04but it's been fantastic chatting with you.
- 15:06SPEAKER_02: Thank you so much for making the time to speak with me.
- 15:07SPEAKER_00: Yeah, no, thank you.
- 15:08SPEAKER_00: It was a pleasure.
- 15:09SPEAKER_00: Appreciate it.
- 15:09SPEAKER_00: Great.
- 15:09SPEAKER_02: Thank you.
- 15:10SPEAKER_02: Thank you to Scott for speaking with us and for sharing those insights.
- 15:15SPEAKER_02: Up next,
- 15:16SPEAKER_02: we're speaking with Muhi Mazoub,
- 15:17SPEAKER_02: EVP of Cybersecurity Engineering at OpenText,
- 15:21SPEAKER_02: about the evolving landscape of cybersecurity threats.
- 15:25SPEAKER_02: Thank you so much for joining me today.
- 15:26SPEAKER_02: It's really great to get a chance to speak with you while I'm here.
- 15:29I'd like to start off by asking you a bit about your background
- 15:34and how you started
- 15:35SPEAKER_02: in cybersecurity,
- 15:36SPEAKER_02: how you came into your current role at OpenText.
- 15:39SPEAKER_03: So I've been an engineer and a software developer for 30 plus years,
- 15:45SPEAKER_03: 17 years at Oracle,
- 15:47SPEAKER_03: six and a half at CA Technology,
- 15:51SPEAKER_03: two years in the UK at Northgate-Orinso outside of London,
- 15:57SPEAKER_03: and then 13 years and a half at OpenTex.
- 16:00In January of this year,
- 16:02I took a new role from,
- 16:04I was for 13 years,
- 16:06the chief product officer.
- 16:08SPEAKER_01: Mm-hmm.
- 16:08SPEAKER_03: I took a new role to build our cybersecurity business unit.
- 16:13And I now manage the three divisions in cybersecurity that does
- 16:18enterprise R&D,
- 16:20SPEAKER_03: SMB and consumer,
- 16:22SPEAKER_03: and then the services division.
- 16:24SPEAKER_02: Excellent.
- 16:25SPEAKER_02: So as the person kind of defining OpenText security product strategy,
- 16:30what areas of security do you believe are most in need of
- 16:34innovation over the next
- 16:35SPEAKER_02: few years?
- 16:35SPEAKER_03: Yeah, that's a great question.
- 16:37SPEAKER_03: And I'll break them into the different areas.
- 16:41SPEAKER_03: In the enterprise space,
- 16:42SPEAKER_03: identity management is one of the most important,
- 16:46SPEAKER_03: followed by application security.
- 16:49followed by data security to secure the data that AI agents are
- 16:55producing.
- 16:56The fourth area will be SIM operation for security operations
- 17:02center and protection
- 17:03SPEAKER_03: against threat,
- 17:04SPEAKER_03: the ability to do threat detection and response.
- 17:08SPEAKER_03: And the fifth area is forensic intelligence and forensic analytics.
- 17:14SPEAKER_02: Excellent.
- 17:15SPEAKER_02: So the cybersecurity space is very fast paced.
- 17:18SPEAKER_02: So when they need to keep up with kind of evolving threats,
- 17:22how can product development and engineering teams balance the
- 17:25need to innovate
- 17:26SPEAKER_02: quickly with the need to ensure governance and security testing?
- 17:30SPEAKER_03: That's a great question.
- 17:31SPEAKER_03: Thank you.
- 17:33SPEAKER_03: So we use AI internally, but we use AI internally.
- 17:39SPEAKER_03: with human in control.
- 17:41So AI is help making our engineers and our managers more
- 17:46efficient by automating
- 17:48tasks that in the past would take three or four hours that we
- 17:52could do it in 15
- 17:53SPEAKER_03: minutes.
- 17:54SPEAKER_03: At the end,
- 17:55SPEAKER_03: an engineer,
- 17:56a human being need to review and validate and confirm that this
- 18:02code is ready.
- 18:03The second level we do,
- 18:05we use our own platform called application security.
- 18:10SPEAKER_03: to test our application for vulnerability,
- 18:14SPEAKER_03: for validating data accuracy,
- 18:18SPEAKER_03: or validating the ethnicity of libraries.
- 18:22SPEAKER_03: An example,
- 18:23an engineer is building a Python or a Java code line could go
- 18:28on the internet and
- 18:29SPEAKER_03: grab 50 LIPR.
- 18:30SPEAKER_01: Mm-hmm.
- 18:31SPEAKER_03: Each of those libraries is thousands of lines of code.
- 18:35SPEAKER_03: They plug them in, they start using them.
- 18:38SPEAKER_03: Our platform comes in and crawl those libraries,
- 18:42crawl the code behind these libraries and identify security
- 18:46vulnerability,
- 18:48SPEAKER_03: open source vulnerability,
- 18:50SPEAKER_03: risks in patch level.
- 18:52SPEAKER_03: So if the current Python library is 19.x and you're using 17.x,
- 18:59You're alerted that you're running two levels behind and you
- 19:03may have security
- 19:05SPEAKER_03: risks because of that.
- 19:07SPEAKER_02: Right, I see.
- 19:07SPEAKER_02: That sounds very useful.
- 19:09What are the main challenges that your team are facing when
- 19:12building products for
- 19:13SPEAKER_02: modern SOC teams?
- 19:15How many connectors can we develop and how fast we can develop
- 19:18them is real.
- 19:19We have a great product that runs in the public cloud that
- 19:24Microsoft asked us to
- 19:25SPEAKER_03: partner with them and to help support two of Microsoft's platform,
- 19:30SPEAKER_03: IntraID and Defender.
- 19:33SPEAKER_03: And they're now asking us to also support their Sentinel platform.
- 19:37SPEAKER_03: So we would like to develop hundreds of connectors.
- 19:41SPEAKER_03: We would like connectors to connect into our own open text platform.
- 19:46SPEAKER_03: We would like connectors to have the ability to connect to SAP,
- 19:50SPEAKER_03: to Salesforce,
- 19:52SPEAKER_03: to other,
- 19:53and to be able to identify vulnerabilities or anomalies when
- 19:58there are security
- 19:59SPEAKER_03: risks.
- 20:00SPEAKER_03: So really,
- 20:00the biggest challenge is finding enough hours in a day to how
- 20:05quickly you could
- 20:06develop these connectors to different systems that behave and
- 20:10operate differently.
- 20:12SPEAKER_02: Right, I see.
- 20:13SPEAKER_02: Following on from that, how do you overcome those challenges?
- 20:17AI helps us become more efficient in speeding up work and then
- 20:22leveraging our own
- 20:23SPEAKER_03: platform,
- 20:24SPEAKER_03: application security,
- 20:26SPEAKER_03: and other solutions that we have in our DevSec.
- 20:30Our core discovery software delivery tools helps us speed up
- 20:36the process and the
- 20:37SPEAKER_03: lifecycle of software development.
- 20:39SPEAKER_02: Right.
- 20:40So how important would you say user feedback is when it comes
- 20:44to delivering new products?
- 20:46How do you access that feedback and translate it into a new
- 20:51product or feature?
- 20:52SPEAKER_03: That's a great question and one of my favorite, actually.
- 20:55SPEAKER_03: And I'll tell you why.
- 20:57SPEAKER_03: We take pride into working very closely with our customers.
- 21:01SPEAKER_03: From them, we learn.
- 21:03The gentleman that sat with me from BMW,
- 21:06Rene,
- 21:06has become a friend over the past three years.
- 21:09SPEAKER_03: I've visited him twice.
- 21:11SPEAKER_03: I've toured his QA processes and life cycle in detail.
- 21:17SPEAKER_03: And when I say I tour,
- 21:19SPEAKER_03: he meant walking through a lab ten times the size of this room,
- 21:24SPEAKER_03: looking at BMW vehicle.
- 21:26SPEAKER_03: electronically getting assembled when they didn't look like cards.
- 21:30They looked like a rack of shelves and they had different
- 21:34modules and connected to a computer.
- 21:37We take pride into collecting that feedback and bringing it
- 21:40back to our development center.
- 21:42SPEAKER_03: Because depending on the industry you work in,
- 21:45SPEAKER_03: unless our customers are in the software industry,
- 21:49SPEAKER_03: if they're in automotive,
- 21:51SPEAKER_03: the customer is smarter than we are.
- 21:53SPEAKER_03: They know their automotive business better than I do.
- 21:57SPEAKER_03: better than my product managers do.
- 21:59SPEAKER_03: So we collect that data.
- 22:01SPEAKER_03: In events like this,
- 22:02SPEAKER_03: we hold customer advisory board where we bring 10,
- 22:07SPEAKER_03: 12,
- 22:1015 customers in a room for half a day.
- 22:10SPEAKER_03: We brief them on our vision feedback.
- 22:13SPEAKER_03: Sometimes we ask them to vote on new feature.
- 22:16We're thinking of doing A,
- 22:18B,
- 22:18C,
- 22:18D,
- 22:19E,
- 22:19F,
- 22:19G,
- 22:20but we can only do three things.
- 22:22SPEAKER_03: Which one would you like us to do first?
- 22:24SPEAKER_03: The customer.
- 22:25will tell you they love it when they give you feedback and they
- 22:29see the feedback as
- 22:31SPEAKER_03: a feature in the product because it means they were heard,
- 22:34SPEAKER_03: listened to them,
- 22:36SPEAKER_03: and we acted.
- 22:37SPEAKER_03: And so we do a lot of that.
- 22:40SPEAKER_03: On a weekly basis, I have an average of five customer meetings.
- 22:44SPEAKER_03: Sometimes there are challenges in technology, challenges in deployment.
- 22:48We take all of that feedback and that helps us enhance our
- 22:52background processes,
- 22:54SPEAKER_03: our automation,
- 22:55SPEAKER_03: and what we deliver in the next release of a product.
- 22:59SPEAKER_02: So this year at OpenTex, the big topic has been AI.
- 23:04So what role does agentic AI play in shaping the future of
- 23:09cybersecurity products?
- 23:11SPEAKER_03: AI plays a big role in cybersecurity,
- 23:14SPEAKER_03: and it's a double-edged sword because it's both offense and defense.
- 23:19It helped the attackers be smarter,
- 23:22but it also helped the defenders,
- 23:24us,
- 23:25be a lot smarter.
- 23:26SPEAKER_03: We hope that we can always stay one step ahead of our attackers.
- 23:32SPEAKER_03: Throughout the last 20 years,
- 23:34we have used AI and machine learning in their earlier stages of
- 23:39technology.
- 23:40Our WebRoot antivirus and endpoint protection product has had
- 23:46machine learning and
- 23:47SPEAKER_03: AI embedded in the engine.
- 23:50SPEAKER_03: for 17 years.
- 23:52SPEAKER_03: So we're proud of that.
- 23:54SPEAKER_03: But we continue to build.
- 23:56SPEAKER_03: Gen AI is something new.
- 23:58SPEAKER_03: We are looking at how best we leverage it.
- 24:01SPEAKER_03: But currently, most of our AI in cybersecurity is embedded in our engine.
- 24:07SPEAKER_03: So we know what we put in it and we don't allow any external influence.
- 24:12What are some of the risks or limitations of AI that you're
- 24:14most mindful of when
- 24:15SPEAKER_02: you're within engineering teams?
- 24:21SPEAKER_03: That's a very good question and a question that could take three hours.
- 24:25SPEAKER_03: The risk is really making the attacker stronger.
- 24:29SPEAKER_03: That's the biggest risk.
- 24:30Prompt injection,
- 24:32I think,
- 24:32will be the next vector of attack in cybersecurity.
- 24:37SPEAKER_03: Hackers will take advantage of being able to manipulate prompt to act.
- 24:43SPEAKER_03: I suspect you will start seeing phishing attack that
- 24:48involve AI as well,
- 24:50because they can mimic voices,
- 24:51they can mimic facial.
- 24:54And so they will make it harder and harder to the point where
- 24:58now internally,
- 25:00I want to encourage our family that we should have a secret
- 25:03code that only our
- 25:05SPEAKER_03: family knows.
- 25:06SPEAKER_03: Anybody text me and say, dad, I need $10,000.
- 25:10SPEAKER_03: I am in a car accident or something.
- 25:13SPEAKER_03: I'll ask you for the code.
- 25:15SPEAKER_03: And hopefully it's something that means something to our family.
- 25:19SPEAKER_03: It may be a word.
- 25:20SPEAKER_03: It may be a word plus numeric.
- 25:23SPEAKER_03: It could be anything.
- 25:25SPEAKER_03: But I think we need to protect ourselves.
- 25:28SPEAKER_03: Those are the risks.
- 25:29SPEAKER_03: On the other side, the efficiencies are too many to name.
- 25:33SPEAKER_03: Too many.
- 25:34SPEAKER_03: I mean, we are already benefiting ourselves.
- 25:37From doing things,
- 25:38leveraging AI publicly and privately inside our firewall.
- 25:43SPEAKER_03: This is a true story in CBS News.
- 25:47SPEAKER_03: An elderly lady lost, I think, $420,000 to scammers.
- 25:55SPEAKER_03: who pretended to be her grandson,
- 25:57SPEAKER_03: who is in college,
- 25:59SPEAKER_03: saying,
- 25:59SPEAKER_03: Grandma,
- 25:59SPEAKER_03: I was in a car crash or I'm in jail.
- 26:05SPEAKER_03: I forgot the detail of the story.
- 26:08SPEAKER_03: But if you Google CBS phishing attack or scammer attack on elderly,
- 26:16SPEAKER_03: you will see a whole list of them.
- 26:18SPEAKER_03: This one, $420,000 in one weekend.
- 26:22SPEAKER_03: They somehow managed through the attack.
- 26:25SPEAKER_03: I'm sending you a page.
- 26:27SPEAKER_03: Please link in.
- 26:28SPEAKER_03: Put your something in it, your credit card.
- 26:31But somehow they got access to her bank account and they
- 26:35cleaned her bank account.
- 26:36SPEAKER_03: It was like overdraft.
- 26:38SPEAKER_03: Yeah.
- 26:39SPEAKER_03: And this is one of many examples.
- 26:42SPEAKER_02: Preys on people's fear and worry for their family and their loved ones.
- 26:46SPEAKER_03: Or lack of knowledge of how to protect themselves.
- 26:49SPEAKER_02: Final question,
- 26:50SPEAKER_02: taking a step back,
- 26:51what should organizations' top cybersecurity planning
- 26:55priorities be going into
- 26:57SPEAKER_02: 2026?
- 26:59SPEAKER_03: That's a great question.
- 27:00I think one and most important is protecting the identity of
- 27:04their employees and
- 27:06SPEAKER_03: any external entities that work with them and has access to system.
- 27:11SPEAKER_03: Protecting identity is top of mind for every CIO.
- 27:15If you ask Shannon Bell,
- 27:17I think that would be the first thing she told you.
- 27:20SPEAKER_03: She put our platform on every mobile device, every iPad, every desktop.
- 27:25SPEAKER_03: Nothing happened unless you authenticate.
- 27:28SPEAKER_03: And you can authenticate with fingerprint if your laptop has it,
- 27:33SPEAKER_03: facial if you're on a mobile device or iPad,
- 27:36SPEAKER_03: or two factor authentication code.
- 27:38SPEAKER_03: That's one.
- 27:39SPEAKER_03: Second, ensuring you have a documented process.
- 27:43that every engineer and every technologist in your company is
- 27:48following in how they
- 27:50SPEAKER_03: develop and roll out applications.
- 27:52SPEAKER_03: What are your standards?
- 27:54SPEAKER_03: How do you develop code?
- 27:55SPEAKER_03: How do you secure the code?
- 27:57SPEAKER_03: How do you validate that that security is good?
- 28:00SPEAKER_01: Mm-hmm.
- 28:01SPEAKER_03: Third, protect the data.
- 28:03SPEAKER_03: Data is critical.
- 28:04SPEAKER_03: It's the currency of a digital world is everything.
- 28:09SPEAKER_03: I don't use money.
- 28:10SPEAKER_03: Everything is digital for me.
- 28:12SPEAKER_03: Credit card is here.
- 28:14SPEAKER_03: Driver license is here.
- 28:16SPEAKER_03: One day the passport will be here.
- 28:18SPEAKER_03: That's all you need.
- 28:19SPEAKER_03: That's the future, isn't it?
- 28:20SPEAKER_03: Yeah.
- 28:20SPEAKER_03: I mean, I've been on the road now.
- 28:22SPEAKER_03: This is day number four.
- 28:24SPEAKER_03: I honestly have not touched my wallet and not touched my cash.
- 28:28SPEAKER_03: Everything on the mobile device.
- 28:30SPEAKER_03: This Amazon order online is everything.
- 28:33Protecting the data becomes critical in ensuring that it
- 28:37doesn't fall in the wrong hands.
- 28:39SPEAKER_02: Absolutely essential.
- 28:40SPEAKER_02: That's all my questions.
- 28:42SPEAKER_03: Thank you for your time.
- 28:43SPEAKER_03: It's a pleasure meeting you.
- 28:44SPEAKER_02: Thank you, Timuhee, for sharing that with us.
- 28:47SPEAKER_02: That's all for this episode of Expert Insights Podcast.
- 28:51Be sure to hit subscribe so that you're the first to hear each
- 28:53episode as soon as it's released.