Latest / Expert Insights Podcast / #69 - SecOps in the Age of AI and Non-Human Identities (Chas Clawson, Sumo Logic)
Transcript
- 0:03SPEAKER_02: Hello and welcome to the Expert Insights Podcast,
- 0:05SPEAKER_02: your go-to source for cybersecurity insights from industry experts.
- 0:09SPEAKER_02: I'm your host, Joel Wits.
- 0:14SPEAKER_02: How will AI redefine the future of security operations?
- 0:18In today's episode,
- 0:19we explore this question with Chas Clawson,
- 0:21the field CTO at Sumo Logic.
- 0:23SPEAKER_02: We talked about how AI is reshaping cybersecurity operations,
- 0:27SPEAKER_02: the promise of agentic AI,
- 0:29and the responsible adoption of AI and its transformative
- 0:32potential in the evolving
- 0:33SPEAKER_02: threat landscape.
- 0:34SPEAKER_02: Let's kick it off.
- 0:37So it'd be great to just kick off a little bit of an
- 0:38introduction to yourself and
- 0:39SPEAKER_03: your background and your role at Sumo Logic.
- 0:41SPEAKER_01: Yeah, so I'm Chas Clausen.
- 0:43SPEAKER_01: I'm the field CTO over security at Sumo Logic.
- 0:46SPEAKER_01: I've been here for about six years.
- 0:48My background is really security operations,
- 0:50so I spent a lot of time building security teams.
- 0:54SPEAKER_01: A lot of it actually in the federal government space.
- 0:56I was at the NSA on their red team for a little bit,
- 0:59so I've seen a lot of interesting things.
- 1:01But yeah,
- 1:01it's a crazy time,
- 1:02as you were saying,
- 1:03to be in this space with the age of AI.
- 1:06SPEAKER_03: Yeah, absolutely.
- 1:07SPEAKER_03: I really want to dig into that and talk about Agentic.
- 1:10But it'd be great to start with the announcements that Sumo
- 1:13Logic have made this
- 1:14SPEAKER_03: week here at the RSA conference.
- 1:16SPEAKER_03: So the new capabilities around detection speed,
- 1:18SPEAKER_03: accuracy,
- 1:19SPEAKER_03: and efficiency,
- 1:19SPEAKER_03: it'd be great to just cover off those.
- 1:21SPEAKER_01: Yeah, for sure.
- 1:21So what we're looking to do with Sumo Logic is a cloud SaaS
- 1:25analytics platform.
- 1:26SPEAKER_01: And so we're trying to leverage the power of these intelligent agents,
- 1:30SPEAKER_01: for example,
- 1:31SPEAKER_01: to chew through a lot of this data and provide actionable insights.
- 1:34SPEAKER_01: At the end of the day,
- 1:35SPEAKER_01: all systems that are built,
- 1:37SPEAKER_01: whether on-prem or in the cloud,
- 1:38SPEAKER_01: have a digital exhaust that's typically in the form of log data,
- 1:42SPEAKER_01: right?
- 1:42SPEAKER_01: So what are the things doing?
- 1:43SPEAKER_01: Are they doing it well?
- 1:44SPEAKER_01: What's the performance?
- 1:46So we consume all of that log data and try and provide it back
- 1:48to our customers in
- 1:50SPEAKER_01: a form that's digestible and actionable.
- 1:52SPEAKER_01: And AI is just one component of that.
- 1:54SPEAKER_01: For example, our co-pilot component, which we call MoPilot for Sumo.
- 1:59SPEAKER_01: So MoPilot really is all about,
- 2:01SPEAKER_01: so there's so much data,
- 2:03but does everybody necessarily have to be a data scientist to
- 2:06make that accessible?
- 2:07SPEAKER_01: Right.
- 2:07SPEAKER_01: Do you have to be a ninja at writing queries?
- 2:09And so the ability to interface with the query language through
- 2:13natural language,
- 2:14SPEAKER_01: hey,
- 2:14SPEAKER_01: show me all the failed logins over the last 24 hours on a map,
- 2:18SPEAKER_01: you know,
- 2:18SPEAKER_01: and go.
- 2:19SPEAKER_01: And it just builds that data for you and that visualization for you.
- 2:22SPEAKER_01: It's really changing the game, right?
- 2:25Because the friction is reduced and consumers can or anybody
- 2:29can just get in and
- 2:29SPEAKER_01: say,
- 2:30SPEAKER_01: hey,
- 2:30SPEAKER_01: I have a problem.
- 2:31SPEAKER_01: I'm going to ask Sumo Logic to see if it has the telemetry.
- 2:34SPEAKER_01: And what's the answer?
- 2:35I think agentic AI has been the kind of number one buzzword at
- 2:39the conference.
- 2:39I think it'd be great to dig into kind of what does it mean for
- 2:42security teams?
- 2:43SPEAKER_03: Why is it so hot in this space and kind of what the impact it's having?
- 2:45SPEAKER_03: And I guess it kind of touches on some of the points we made there.
- 2:48SPEAKER_01: Yeah.
- 2:48SPEAKER_01: So agentic AI,
- 2:50SPEAKER_01: I mean,
- 2:50this is kind of the second year that that was like really
- 2:53gathering momentum.
- 2:54It's interesting because the industry hasn't really coalesced
- 2:57around what actually that means.
- 2:59SPEAKER_01: What's the definition of agentic AI?
- 3:01And in my opinion,
- 3:02there's two schools of thoughts,
- 3:03and they're both,
- 3:04I would say,
- 3:04accurate.
- 3:05On one hand,
- 3:06it's goal-oriented,
- 3:07autonomous agents that are doing something,
- 3:10right?
- 3:10SPEAKER_01: Automating something for you in an intelligent way.
- 3:13The other flavor of agentic AI is multiple agents that are
- 3:17working together in a
- 3:18SPEAKER_01: coordinated fashion,
- 3:19SPEAKER_01: right?
- 3:19SPEAKER_01: And maybe there's a master agent or whatever the case may be,
- 3:23SPEAKER_01: but you break apart a large problem down into...
- 3:27SPEAKER_01: palatable areas and then you attach an agent to each one of those.
- 3:30And so,
- 3:31again,
- 3:31I think they're both valid as far as security is concerned.
- 3:34I think we're going to see it embedded in every aspect of the
- 3:38investigation and the
- 3:39SPEAKER_01: response life cycle.
- 3:41SPEAKER_01: So like from beginning to end, if you look at...
- 3:44Like as it's evolving today or maybe yesterday,
- 3:47it was human led investigations,
- 3:50AI assisted.
- 3:51SPEAKER_01: Now,
- 3:51SPEAKER_01: I would say tomorrow or whenever tomorrow is could be literally tomorrow,
- 3:55but it's going to be AI led and human reviewed where the AI
- 3:59does the bulk of the
- 4:00SPEAKER_01: heavy lifting,
- 4:01SPEAKER_01: the correlation,
- 4:02SPEAKER_01: the.
- 4:03you know,
- 4:03alerting and then the human comes over top and says,
- 4:06can I validate this?
- 4:08Do I want to take some autonomous response based on what this
- 4:11agent is telling me to do?
- 4:12And that changes the game because now all of a sudden your
- 4:16analyst can be more efficient.
- 4:17SPEAKER_01: You can operate with less analysts.
- 4:19SPEAKER_01: You can automate so much of the tedious workflows that have,
- 4:23SPEAKER_01: you know,
- 4:23SPEAKER_01: always been challenging,
- 4:25like alert fatigue and swivel chair syndrome is another phrase
- 4:29that they used to
- 4:29SPEAKER_01: use.
- 4:30SPEAKER_01: All of that can be done by AI.
- 4:31SPEAKER_01: So it's interesting.
- 4:33SPEAKER_03: And how far along do you think we are in the agency journey?
- 4:38SPEAKER_03: In terms of security, I suppose, the bigger picture.
- 4:41SPEAKER_03: Are organizations really using these tools today?
- 4:44So usually I'm the guy in the back of the room rolling my eyes
- 4:47and be like,
- 4:48SPEAKER_01: AI,
- 4:49SPEAKER_01: ML,
- 4:49SPEAKER_01: or Zero Trust,
- 4:51SPEAKER_01: or blockchain,
- 4:52SPEAKER_01: or whatever technology.
- 4:53SPEAKER_01: And it's like...
- 4:54SPEAKER_01: Yeah, I get it.
- 4:55SPEAKER_01: But to be honest, I think this is different, right?
- 4:58And I think it's primarily because the broad capabilities that
- 5:03this brings and the
- 5:04SPEAKER_01: way that it can be used in so many different areas is,
- 5:08SPEAKER_01: I mean,
- 5:09SPEAKER_01: we've barely broken the surface,
- 5:10SPEAKER_01: right?
- 5:11So GPT launched,
- 5:12that GPT launched,
- 5:13you know,
- 5:14a couple years ago,
- 5:15a few years ago at most.
- 5:16SPEAKER_01: And
- 5:18SPEAKER_01: The rate of adoption is crazy.
- 5:21SPEAKER_01: And the intensity in which you used to use like GPUs and power and,
- 5:28SPEAKER_01: you know,
- 5:29SPEAKER_01: the hardware infrastructure used to be unattainable.
- 5:32SPEAKER_01: We're seeing that drop dramatically,
- 5:34which means now businesses that previously couldn't afford to
- 5:37take this authentic
- 5:38SPEAKER_01: AI capability can now afford to do it.
- 5:41SPEAKER_01: right and so what i expect to see happening is the costs drop off a cliff
- 5:46everything gets more efficient and then we start to see it
- 5:49embedded in areas that
- 5:50we didn't even conceive of right from iot devices at home your
- 5:54phone your glasses
- 5:55your you know everything is going to have some purpose-built
- 5:59agent as part of its
- 6:00it's just going to be commonplace right yeah and i think
- 6:03literally we're on the
- 6:04SPEAKER_01: verge of just seeing technology explode in very interesting ways
- 6:10You're listening to the Expert Insights Podcast,
- 6:13home of leading cybersecurity conversations.
- 6:16Subscribe now to get insights from top cybersecurity leaders
- 6:19straight to your inbox.
- 6:21And we're already kind of hearing or at least I've heard a lot
- 6:24of stories this week
- 6:24SPEAKER_03: about how CEOs,
- 6:25SPEAKER_03: you know,
- 6:25SPEAKER_03: they don't want to be left behind.
- 6:27Businesses want to adopt these tools quickly and they're kind
- 6:29of compromising those
- 6:30SPEAKER_03: guardrails for AI adoption.
- 6:34SPEAKER_01: Yeah.
- 6:35SPEAKER_01: So it is an interesting challenge.
- 6:38If you look at a business from like a high level,
- 6:41there's two forms of risk.
- 6:42SPEAKER_01: One would be like cyber risk.
- 6:44SPEAKER_01: So like you want to move fast and break things,
- 6:47SPEAKER_01: but you want to do it in a responsible way that doesn't expose,
- 6:50SPEAKER_01: you know,
- 6:50SPEAKER_01: sensitive data or,
- 6:52SPEAKER_01: you know,
- 6:53SPEAKER_01: vulnerabilities.
- 6:53SPEAKER_01: But there's an entirely different risk, which is the innovation risk.
- 6:57SPEAKER_01: And that really is if you're not moving fast enough,
- 7:00SPEAKER_01: your business will wither and you will be usurped by the other guys.
- 7:04SPEAKER_01: We saw this play out with cloud in a much slower scale.
- 7:07SPEAKER_01: But if you don't adopt cloud quickly, then you're going to be less agile.
- 7:12SPEAKER_01: Your velocity is going to be slower than the other guy.
- 7:15And so the businesses that are left standing today are all in
- 7:18on whatever cloud
- 7:19SPEAKER_01: infrastructure is code,
- 7:20SPEAKER_01: et cetera.
- 7:21SPEAKER_01: Well,
- 7:21take that and condense that down to even a more accelerated
- 7:25cycle where if you're
- 7:27SPEAKER_01: not adopting AI today,
- 7:28SPEAKER_01: expect to be swallowed up.
- 7:30SPEAKER_01: And so that's really the challenge,
- 7:32especially for Sumo Logic customers is we're trying to enable
- 7:35them to be aggressive
- 7:37SPEAKER_01: with their AI strategy,
- 7:38SPEAKER_01: but also wrap around it controls,
- 7:40SPEAKER_01: wrap it around monitoring,
- 7:42SPEAKER_01: observability,
- 7:43things that make them a little bit safer because you it's just
- 7:48a balancing act.
- 7:48SPEAKER_01: Right.
- 7:49SPEAKER_01: And that's that's the name of the game.
- 7:51SPEAKER_03: And I suppose with Argentic, it gets even harder.
- 7:53SPEAKER_03: They're talking for each other.
- 7:54SPEAKER_03: It's harder to control what happens.
- 7:56SPEAKER_03: What's the importance of having,
- 7:57SPEAKER_03: I guess,
- 7:58SPEAKER_03: like a human in the loop or at least some mechanism to control that?
- 8:01SPEAKER_01: Yeah,
- 8:01SPEAKER_01: so the non-human identity problem,
- 8:04SPEAKER_01: which maybe you're familiar with,
- 8:05SPEAKER_01: which is like,
- 8:06SPEAKER_01: you know,
- 8:07SPEAKER_01: in a typical enterprise,
- 8:09SPEAKER_01: non-human identities outnumber humans 10 to 1.
- 8:11SPEAKER_01: This is the number that I hear.
- 8:13SPEAKER_01: And so you're typically monitoring humans closely,
- 8:16SPEAKER_01: like what is Bob doing or Mary doing to check his behaviors,
- 8:20SPEAKER_01: etc.
- 8:21But these 10 other service accounts or non-human entities don't
- 8:25get as much attention.
- 8:26Imagine taking these service accounts,
- 8:28which are benign,
- 8:29injecting them with intelligence.
- 8:31SPEAKER_01: Now they're autonomous.
- 8:32SPEAKER_01: They can actually take action.
- 8:33SPEAKER_01: They could do things.
- 8:35SPEAKER_01: They could be goal oriented.
- 8:36SPEAKER_01: Right.
- 8:37SPEAKER_01: And so to your point,
- 8:38are we going to have an explosion of non-human identities in
- 8:42all walks of life and
- 8:43SPEAKER_01: all of our systems?
- 8:44SPEAKER_01: And if that's the case, are we monitoring what's happening?
- 8:47SPEAKER_01: Right.
- 8:47SPEAKER_01: If something goes awry, do we have the telemetry to identify it quickly?
- 8:52SPEAKER_01: and bring it back to ground.
- 8:53And I saw an interesting article just last week where one of
- 8:57the chatbots responded
- 8:58SPEAKER_01: to a customer and it was through,
- 9:00SPEAKER_01: I think,
- 9:00SPEAKER_01: a web interface.
- 9:01SPEAKER_01: And it said,
- 9:02SPEAKER_01: well,
- 9:03company's policy is X and therefore you don't get a refund or
- 9:06it was something like
- 9:07SPEAKER_01: that.
- 9:07SPEAKER_01: Right.
- 9:08SPEAKER_01: And then all of a sudden the company is legally bound to like,
- 9:12SPEAKER_01: hey,
- 9:12SPEAKER_01: this chatbot was speaking on your behalf.
- 9:14Whether that is actually your policy doesn't matter because the
- 9:18customer,
- 9:18you are that agent.
- 9:20And so did they have proper controls monitoring the inputs and
- 9:23the outputs?
- 9:24SPEAKER_01: And, you know, were they logging that interaction?
- 9:26SPEAKER_01: Would they have seen it?
- 9:27SPEAKER_01: How bad could that have gotten?
- 9:28SPEAKER_01: You know, so that was a one-off edge case.
- 9:31SPEAKER_03: But it does happen.
- 9:32SPEAKER_03: These agents, they're not infallible.
- 9:34SPEAKER_03: They can be tricked and always very difficult to do it.
- 9:37SPEAKER_01: Yeah.
- 9:38SPEAKER_01: Yeah.
- 9:38And when you're basing like an entire business model on some
- 9:41sort of autonomous AI
- 9:42SPEAKER_01: agent,
- 9:43SPEAKER_01: you have to,
- 9:43SPEAKER_01: I mean,
- 9:44SPEAKER_01: just like you would a real employee,
- 9:45SPEAKER_01: you have to make sure they're following policy.
- 9:47SPEAKER_01: You have to have governance.
- 9:49And that's one of the things Sumo Logic does really well is we
- 9:51don't care what kind
- 9:52SPEAKER_01: of data it is.
- 9:53SPEAKER_01: Bring it in and let us do the analytics and the alerting for you.
- 9:56SPEAKER_01: So if something goes off the rails, hey, here's an alert, respond to it.
- 10:00And I think I want to talk a little bit about how to make
- 10:03agency AI deployments
- 10:04SPEAKER_03: successful and why,
- 10:06SPEAKER_03: you know,
- 10:07having a strong database behind it or a data lake behind it is
- 10:10so important behind
- 10:11SPEAKER_03: us.
- 10:11What happens if the data feeding agents is flawed and kind of
- 10:14how you secure against that?
- 10:16SPEAKER_01: That's a tough one.
- 10:18SPEAKER_01: Data is the currency in which AI operates, right?
- 10:21SPEAKER_01: That's how it's built.
- 10:22SPEAKER_01: That's how these neural nets and like the tokens and everything.
- 10:24SPEAKER_01: I mean, that's building the intelligence.
- 10:26SPEAKER_01: And so by design,
- 10:28they're intended to just consume all data and then parse
- 10:32through that and figure
- 10:32SPEAKER_01: out what's accurate and what's not.
- 10:34SPEAKER_01: And like you said, it doesn't always get it right.
- 10:35SPEAKER_01: It hallucinates.
- 10:36SPEAKER_01: I think on one hand,
- 10:38every company is taking generative AI and LLMs as a
- 10:42sledgehammer to chop down every
- 10:44SPEAKER_01: tree.
- 10:45SPEAKER_01: And it's not necessarily the right approach, right?
- 10:47Because you can take purpose-built models or even traditional
- 10:51machine learning and
- 10:52AI and solve problems very efficiently without going through
- 10:56the generative AI.
- 10:57So the question is like,
- 10:58well,
- 10:58why is everybody leaning into gen AI so much?
- 11:00Because it's easy and the friction is gone and anybody can log
- 11:03in and just do it,
- 11:05right?
- 11:05It's harder to implement the traditional tried and true,
- 11:09you know,
- 11:10approaches.
- 11:11But I think,
- 11:12you know,
- 11:12to the poisoning and the toxicity like that,
- 11:14exactly.
- 11:15I think over time,
- 11:16we'll start to see built-in checks and balances,
- 11:19right?
- 11:19Where LLMs can validate each other and models can cross-check
- 11:23and you can like,
- 11:24SPEAKER_01: all of that can be fleshed out over time through code and,
- 11:26SPEAKER_01: you know,
- 11:27SPEAKER_01: whatever the case may be.
- 11:28But we'll also start to see people reevaluate like what models
- 11:32should be used where
- 11:33SPEAKER_01: instead of this,
- 11:33SPEAKER_01: like,
- 11:34just drop a chat bot and everything right so yeah what uh kind
- 11:39of advice or kind of
- 11:41guardrails would you recommend that security themes put in
- 11:43place today to kind of
- 11:44help cure their ai usage you know there's there's formal um
- 11:49frameworks now which is
- 11:50nice like iso 42001 i think is a good place to start because
- 11:55they've already gone
- 11:55through a lot of the thought process of like what does good ai
- 11:59strategy look like
- 12:00SPEAKER_01: Unfortunately, I'm not seeing that like adopted yet.
- 12:03And mostly because we're in the building mode and not the take
- 12:06a step back and look
- 12:07SPEAKER_01: at what we built.
- 12:08SPEAKER_01: And did we do it safely?
- 12:09But I would say just be,
- 12:11you know,
- 12:11it's trite,
- 12:12but be thoughtful about it.
- 12:14SPEAKER_01: And so everything that you build should have,
- 12:18SPEAKER_01: you know,
- 12:18SPEAKER_01: you should really think about those,
- 12:20SPEAKER_01: especially in cases of sensitive data,
- 12:22SPEAKER_01: right?
- 12:22SPEAKER_01: I'm thinking of healthcare, financial industry, government,
- 12:26SPEAKER_01: There's so much that could go wrong.
- 12:27SPEAKER_01: And oftentimes we're so focused on a solution that we don't think,
- 12:31SPEAKER_01: okay,
- 12:31SPEAKER_01: well,
- 12:31SPEAKER_01: what's the worst case scenario that could happen with this data?
- 12:34Or what's the worst case scenario that could happen if the
- 12:37hallucinations took place?
- 12:39SPEAKER_01: So just be thoughtful about how you're building solutions, I think.
- 12:42And more broadly,
- 12:43are you kind of optimistic,
- 12:45excited,
- 12:46concerned,
- 12:46nervous?
- 12:47SPEAKER_03: How do you sort of feel about it?
- 12:48SPEAKER_01: Oh, man, extremely optimistic.
- 12:50SPEAKER_01: I mean,
- 12:51SPEAKER_01: I think it's unlocked,
- 12:52SPEAKER_01: like,
- 12:53SPEAKER_01: just knowledge that was previously unattainable down to the lowest level.
- 12:57SPEAKER_01: I mean, I've got five kids and...
- 12:59SPEAKER_01: You know,
- 13:00when I log into my chat GPT account and I say and I see like
- 13:04you've run out of
- 13:05SPEAKER_01: query quotas for the day and I'm like,
- 13:07SPEAKER_01: I haven't even used it.
- 13:08And I'm looking at the history and it's like solving math
- 13:12problems for my 10 year old.
- 13:14And I'm like,
- 13:14oh,
- 13:14my gosh,
- 13:15like,
- 13:15do we really realize what we've built here yet?
- 13:19SPEAKER_01: and I'm not going to go to my kids and say,
- 13:21SPEAKER_01: you shouldn't be using this technology because I think that's the future.
- 13:25SPEAKER_01: Right.
- 13:25SPEAKER_01: And so from the lowest level to the,
- 13:28SPEAKER_01: to the highest level,
- 13:29SPEAKER_01: I think that this is going to change everything.
- 13:32SPEAKER_01: Right.
- 13:32SPEAKER_01: And so, um, it's not really a hype.
- 13:36SPEAKER_01: This is not a hype cycle.
- 13:38SPEAKER_01: This is like,
- 13:39SPEAKER_01: we're all getting run over by this train,
- 13:41SPEAKER_01: whether you like it or not,
- 13:42SPEAKER_01: and just get as ready as you can in a good way,
- 13:45SPEAKER_01: in a good way.
- 13:46SPEAKER_01: That's great.
- 13:47How do you sort of see this evolving over the next sort of 12
- 13:49months or next couple of years?
- 13:51How do you sort of predict this is going to play out in the
- 13:53security space?
- 13:54SPEAKER_01: In the security space, it's tough.
- 13:56Like you walk the floor here at RSA and there's a lot of
- 13:59different innovative
- 14:00SPEAKER_01: solutions in there.
- 14:01A lot of it is quote unquote AI washed where they're slapping
- 14:05the AI on top of it.
- 14:06SPEAKER_01: But a lot of it is actually real.
- 14:09SPEAKER_01: And when you listen to Sam Altman from OpenAI say,
- 14:13SPEAKER_01: we're on the verge of seeing,
- 14:14SPEAKER_01: you know,
- 14:15million or billion dollar companies being built by teams of
- 14:18five or 10 people,
- 14:19SPEAKER_01: that's possible.
- 14:20SPEAKER_01: is because AI helps from ideation to building the business case,
- 14:25SPEAKER_01: building the architectural diagrams to actual code.
- 14:28And I'm working on some research right now,
- 14:30which is it's going to be a talk.
- 14:32It's called,
- 14:33you know,
- 14:33leveraging AI for rapid prototyping of cyber solutions.
- 14:37SPEAKER_01: And I think you have to broaden your, your use case for AI.
- 14:41SPEAKER_01: It's not all about writing code.
- 14:42SPEAKER_01: Start at the very beginning.
- 14:44SPEAKER_01: I help you identify a problem.
- 14:46SPEAKER_01: Help it build a use case.
- 14:48SPEAKER_01: Can it be monetized?
- 14:49SPEAKER_01: Help it build your marketing plan.
- 14:51SPEAKER_01: Help it build the architectural diagram.
- 14:53SPEAKER_01: This is all happening before you write a single line of code.
- 14:56SPEAKER_01: and now you're getting into,
- 14:57SPEAKER_01: okay,
- 14:57SPEAKER_01: now build this in,
- 14:58SPEAKER_01: you know,
- 14:59SPEAKER_01: infrastructure as code,
- 15:00SPEAKER_01: zero trust,
- 15:00SPEAKER_01: whatever.
- 15:01SPEAKER_01: And so you infuse AI at every single step of the way.
- 15:04SPEAKER_01: You can build a solution and,
- 15:06SPEAKER_01: you know,
- 15:06SPEAKER_01: who knows some of these booths down here on the floor may be a very,
- 15:10SPEAKER_01: very small team building very powerful solutions.
- 15:13SPEAKER_01: I'm sure that as that technology evolves, they'll get
- 15:16platform i you know the big players will will cherry pick out
- 15:19what they think is
- 15:20the most compelling solution and wrap it into like a bigger
- 15:23play but but yeah it's
- 15:24interesting to see like what a small team can do with a little
- 15:28bit of you know
- 15:28horsepower provided by ai chas thank you so much yeah with me
- 15:32with us really
- 15:32SPEAKER_03: appreciate it yeah thank you
- 15:45This podcast was produced by Expert Insights,
- 15:47home of leading cybersecurity conversations.
- 15:50Subscribe now to get insights from top cybersecurity leaders
- 15:54straight into your inbox.
- 15:55Why not check out our weekly decrypted cybersecurity briefings
- 15:59and our Game
- 15:59SPEAKER_00: Changers conversations with innovative individuals.