Latest / Expert Insights Podcast / #58 – Inside Google’s View on Evolving Cyber Risks (John Hultquist, Google Cloud)
Transcript
- 0:04SPEAKER_00: What does it take to outsmart today's cybercriminals?
- 0:08SPEAKER_00: In today's episode,
- 0:09SPEAKER_00: we explore this question and more with John Holtquist,
- 0:12SPEAKER_00: the Chief Analyst at Google Threat Intelligence Group.
- 0:15SPEAKER_00: From Google's unique perspective,
- 0:17SPEAKER_00: John unpacks the growing risks of zero-day exploits and insider threats,
- 0:21SPEAKER_00: the critical importance of proactive security investment,
- 0:24and how AI is poised to influence the future of threat
- 0:27detection and defense.
- 0:29SPEAKER_00: Let's get started.
- 0:34It'd be great just to kick off with a kind of introduction to
- 0:36yourself and your
- 0:37SPEAKER_01: kind of security background and your role at Google.
- 0:39SPEAKER_02: Sure.
- 0:40SPEAKER_02: I'm John Holquist.
- 0:41SPEAKER_02: I'm chief analyst at Google Threat Intelligence Group.
- 0:43I've been doing threat intelligence or cyber threat
- 0:46intelligence for almost 20
- 0:50SPEAKER_02: years now or something like that.
- 0:51SPEAKER_02: Maybe 15 years.
- 0:52SPEAKER_02: I don't know.
- 0:53Most of my time is spent talking to customers and dealing with
- 0:59basically discussing
- 1:00SPEAKER_02: the threats that we are seeing as a group.
- 1:02And what have kind of been the headlines that you've been
- 1:04talking about here at RSAC this week?
- 1:06SPEAKER_01: You know, we were at the panel this morning talking about Agentic AI.
- 1:09SPEAKER_01: I know you've had some big announcements in that kind of.
- 1:12SPEAKER_02: You know, I think, you know, the AI thing is really interesting.
- 1:16We've been tracking for several years now the realm of the
- 1:20possible or even the theoretical.
- 1:22SPEAKER_02: And then there's like the practical, right, what we actually see.
- 1:24And so we put out a report pretty recently on what we were
- 1:28seeing adversaries doing.
- 1:29SPEAKER_02: And what they're really using AI for,
- 1:32SPEAKER_02: at least so far,
- 1:32SPEAKER_02: is social engineering and sort of scaling their operations.
- 1:37And to a lesser extent,
- 1:38sort of like a beefed up search capability,
- 1:41right?
- 1:42They've been doing that sort of using it for a social engineer
- 1:44for quite some time
- 1:45SPEAKER_02: because it's so good at sort of creating messages or,
- 1:49SPEAKER_02: you know,
- 1:49SPEAKER_02: like they could tell it to say,
- 1:52SPEAKER_02: make a message to a colleague or something like that.
- 1:54SPEAKER_02: They can sort of create that for you.
- 1:56SPEAKER_02: It's good at sort of fabricating content from,
- 1:59SPEAKER_02: you know,
- 1:59SPEAKER_02: like fake pictures and audio and video.
- 2:05SPEAKER_02: And so mostly that's been applied to the social engineering space and the
- 2:10SPEAKER_02: information operation space.
- 2:12I think,
- 2:13you know,
- 2:13the big advance has been allowing them to sort of scale up
- 2:18these operations.
- 2:19SPEAKER_02: So if you think of like
- 2:20SPEAKER_02: An adversary that's essentially social engineering, right?
- 2:23If you kind of put one person who speaks poor English and they
- 2:26have to slowly sort
- 2:27SPEAKER_02: of converse somebody convincingly,
- 2:29SPEAKER_02: right?
- 2:29SPEAKER_02: That's a huge ask.
- 2:31And so if you can essentially use this to help them host those
- 2:34conversations or
- 2:36SPEAKER_02: have those conversations along periods of time,
- 2:38SPEAKER_02: it could be a real uplift.
- 2:39SPEAKER_02: A long time ago,
- 2:40SPEAKER_02: or not a long time ago,
- 2:41SPEAKER_02: not long ago,
- 2:42SPEAKER_02: a lot of these adversaries really,
- 2:43SPEAKER_02: when they carried out social engineering,
- 2:45SPEAKER_02: they would send you like a link.
- 2:46SPEAKER_02: They would ask you, say something and then send you a link, right?
- 2:49SPEAKER_02: And that was the whole thing.
- 2:50play increasingly we see them essentially targeting somebody in
- 2:53a long-term
- 2:54conversation right and that's a huge like it could be like a
- 2:57month before they ever
- 2:58send you something malicious and that's a huge lift yeah right
- 3:01and this sort of
- 3:01helps with that yeah and it'd be great you know just to talk
- 3:04about your threat
- 3:04intelligence team you know what's the scale of how you're able
- 3:08to look at the
- 3:09SPEAKER_01: threats you must have kind of a unique vantage point yeah
- 3:11SPEAKER_02: The whole team's got a really incredible vantage point.
- 3:14SPEAKER_02: I mean,
- 3:14SPEAKER_02: the first one is we obviously,
- 3:15SPEAKER_02: we can see the threats that Google is,
- 3:17SPEAKER_02: you know,
- 3:18SPEAKER_02: that's sort of coming at Google.
- 3:19SPEAKER_02: Google's a massive attack surface, right?
- 3:22Obviously,
- 3:22we're constantly being targeted by our customers,
- 3:27who we're protecting,
- 3:28our users.
- 3:29SPEAKER_02: And by defending them all, we're learning, right?
- 3:32Really,
- 3:32the name of the game is that every time something sort of bad
- 3:35happens,
- 3:35that you take
- 3:36lesson away from it right and you learn more about the
- 3:39adversary yeah and that's
- 3:40sort of what we're trying to do so we learn about these things
- 3:43these threats from
- 3:44from the things that are targeting our users and our customers
- 3:47we learn about them
- 3:48in the in the sort of dark web we we learn about them through
- 3:51things like virus
- 3:52SPEAKER_02: total you know which was a huge repository of information
- 3:56And the other big place that we learned about a lot of things
- 3:58is the incident
- 3:59SPEAKER_02: response that Mayonet does,
- 4:00SPEAKER_02: right?
- 4:00So Mayonet generally gets the call when something very serious
- 4:04happens at the
- 4:04SPEAKER_02: organization and they don't want to screw around and they call the pest.
- 4:08And that gives us visibility and really the number of the most
- 4:11important incidents.
- 4:12SPEAKER_02: And by aggregating all that information,
- 4:14we can essentially develop a picture of the threat that's more
- 4:18refined than you
- 4:19SPEAKER_02: could get just about anywhere.
- 4:21SPEAKER_02: You have to think that we're essentially tracking criminals and spies,
- 4:25SPEAKER_02: and they're doing everything possible not to let us learn about them,
- 4:28SPEAKER_02: right?
- 4:28And so it takes a lot of work to build out a picture of them,
- 4:33and you have to do that.
- 4:34SPEAKER_02: If you want to do any kind of risk,
- 4:37SPEAKER_02: like risk-based security, which is the only way to do security, right?
- 4:41SPEAKER_02: You have to actually know what's coming at you.
- 4:43SPEAKER_02: We got a picture of a castle behind us here.
- 4:45SPEAKER_02: I think like, you know, a castle, it's great.
- 4:48SPEAKER_02: It's all, well, ironically, it's a castle with helicopters.
- 4:51SPEAKER_02: And they're like...
- 4:53SPEAKER_02: A castle's okay.
- 4:53It might be a great way to defend yourself until the adversary
- 4:56has helicopters.
- 4:57SPEAKER_02: Right.
- 4:58SPEAKER_02: Right.
- 4:58SPEAKER_02: And then it's obsolete.
- 5:00SPEAKER_02: And so you think about like how you're defending yourself, right?
- 5:02SPEAKER_02: You have to understand whether or not the adversary has helicopters.
- 5:06SPEAKER_02: And that's sort of what we're here for.
- 5:07SPEAKER_01: Yeah.
- 5:08And in terms of the threats,
- 5:09you know,
- 5:10you mentioned AI and social engineering.
- 5:11SPEAKER_01: Is that the top threat factor?
- 5:13SPEAKER_02: No,
- 5:13SPEAKER_02: I think that,
- 5:14SPEAKER_02: you know,
- 5:14SPEAKER_02: the things that really worry,
- 5:16SPEAKER_02: I guess,
- 5:16SPEAKER_02: the most is a couple of things.
- 5:18When it comes to adversaries from China and also Russia and
- 5:22just a lot of the more
- 5:23SPEAKER_02: sophisticated or more advanced,
- 5:25SPEAKER_02: technically advanced state actors,
- 5:27SPEAKER_02: we're seeing a lot of pressure on the network edge.
- 5:30SPEAKER_02: You know, Iran and North Korea usually target through social engineering.
- 5:36SPEAKER_02: Russia, China, we're seeing more of their activity on the network edge.
- 5:39They're particularly interested in security devices that sit on
- 5:42the network edge.
- 5:42SPEAKER_02: Right.
- 5:43is they're difficult to monitor and it's an opportunity for
- 5:47them to essentially get
- 5:48a beachhead in the network and then move laterally where you
- 5:52can't necessarily see
- 5:53SPEAKER_02: them or it's really difficult to.
- 5:55That's a huge piece of what we're seeing as far as defending
- 5:59our customers and
- 6:00SPEAKER_02: users against those sorts of actors.
- 6:03SPEAKER_02: The other thing that
- 6:05SPEAKER_02: you know, we're really worried about is the criminal threat, right?
- 6:07I think one of the opportunities or the sort of real drivers of
- 6:12the criminal threat
- 6:12SPEAKER_02: that we've seen lately has been that sort of info stealer space.
- 6:16So info stealers were really huge back in the day of fraud when
- 6:20fraud was like the
- 6:22SPEAKER_02: criminal problem and they kind of fell off.
- 6:24And the reason they were so popular fraud is because they were
- 6:27essentially,
- 6:27they would sit on somebody's machine and then steal other
- 6:30credit card numbers and
- 6:31SPEAKER_02: PII and things like that.
- 6:33And a criminal's job was basically to basically slowly gather
- 6:36all the stuff,
- 6:37right?
- 6:37SPEAKER_02: And then, you know, sell it off for fractions of a set, right?
- 6:42SPEAKER_02: And they would monetize it.
- 6:43SPEAKER_02: It's a really complicated process.
- 6:45SPEAKER_02: If they were lucky,
- 6:46SPEAKER_02: they would hit some,
- 6:47SPEAKER_02: like,
- 6:48SPEAKER_02: bank or retail or something and get all that information at one time,
- 6:51SPEAKER_02: right?
- 6:51SPEAKER_02: That was basically how, like, the criminal space worked at the time.
- 6:55With ransomware and extortion schemes and sort of change the
- 6:59way that a lot of the
- 7:00SPEAKER_02: criminal activity played out,
- 7:02at first they weren't really that interested in infrasilers
- 7:06until they realized
- 7:06SPEAKER_02: there's a great way to get initial access in an organization.
- 7:11So instead of really focusing on using them to gather that PII
- 7:16or the credit card
- 7:16SPEAKER_02: information,
- 7:17they're really using it to gather credentials and sort of like
- 7:21a means to establish
- 7:22SPEAKER_02: information
- 7:23SPEAKER_02: you know,
- 7:23SPEAKER_02: a way into an organization,
- 7:26SPEAKER_02: which might lead to ransomware incentive or something like that.
- 7:28And I think Google has just released a report on zero-day
- 7:32exploits as well,
- 7:32right?
- 7:33SPEAKER_01: What are the kind of key takeaways that you saw from that?
- 7:36SPEAKER_02: So it's good news, bad news.
- 7:38SPEAKER_02: Good news is that we have sort of a lower number than last year,
- 7:43SPEAKER_02: but I think ultimately we're still training upwards,
- 7:48SPEAKER_02: right?
- 7:49SPEAKER_02: And so it might be
- 7:51It might be a little bit,
- 7:52the lower number might be slightly lower number.
- 7:54SPEAKER_02: It's a little higher than it was two years ago.
- 7:56SPEAKER_02: It might be a bit misleading.
- 7:58We're definitely seeing a lot of pressure,
- 8:00for instance,
- 8:01on those edge devices.
- 8:04SPEAKER_02: And a lot of activity from China,
- 8:09SPEAKER_02: which is,
- 8:10SPEAKER_02: it really has industrialized the process of developing these zero days.
- 8:14The other real key takeaway is that North Korea has been
- 8:17involved in quite a few zero days.
- 8:19SPEAKER_02: which is also kind of disconcerting.
- 8:22SPEAKER_02: I'll tell you,
- 8:22SPEAKER_02: you know,
- 8:23the funny thing about like the zero-day problem is that there's
- 8:25a time when we told
- 8:26SPEAKER_02: people like,
- 8:28SPEAKER_02: you're not cool enough to warrant a zero-day,
- 8:31SPEAKER_02: right?
- 8:32SPEAKER_02: They want to,
- 8:33you really want to control the use of a zero-day because if you
- 8:36use it too much,
- 8:37SPEAKER_02: right,
- 8:37SPEAKER_02: it gets out and then it gets fixed.
- 8:39SPEAKER_02: And so you can't just use it everywhere else.
- 8:41And previously,
- 8:42a lot of zero days were used in social engineering type
- 8:46attacks.
- 8:47SPEAKER_02: And so there's a massive risk, right?
- 8:49SPEAKER_02: Like right off the bat, you've basically got to fool somebody.
- 8:52SPEAKER_02: And if they see through your ruse,
- 8:55SPEAKER_02: right,
- 8:55you could lose access to that zero day and it can get like
- 8:58patched and the value
- 8:59SPEAKER_02: dissipates overnight.
- 9:00But because a lot of these zero days are sort of edge-faced and
- 9:05they're basically
- 9:06SPEAKER_02: hitting a device that doesn't have great visibility,
- 9:09SPEAKER_02: they can be used for a lot more targets.
- 9:12And so I'm increasingly not telling people that they're not
- 9:16cool enough because
- 9:17SPEAKER_02: they may well be cool enough for a zero day now,
- 9:20SPEAKER_02: right?
- 9:20Because they're leveraging these zero days far,
- 9:23far more than they did in the past.
- 9:25SPEAKER_01: And you mentioned North Korea there and the Google event on Monday.
- 9:29I was trying to,
- 9:30I think I'm re-sharing quite a bit about the fake worker scam.
- 9:34It'd be great to hear kind of from your end what you sort of
- 9:37see with that attack
- 9:37SPEAKER_02: and what kind of...
- 9:41SPEAKER_02: It's been highly country in the United States,
- 9:43SPEAKER_02: but we've seen in the UK,
- 9:44SPEAKER_02: for instance,
- 9:45SPEAKER_02: in Europe now.
- 9:45And one of our concerns is that they're under a lot of pressure
- 9:48in the United
- 9:48SPEAKER_02: States,
- 9:49SPEAKER_02: particularly with some law enforcement actions,
- 9:50that they're probably increasingly moving it to places where
- 9:53they don't have the
- 9:53SPEAKER_02: same experience yet.
- 9:54SPEAKER_02: Right.
- 9:55SPEAKER_02: But it's pretty wild.
- 9:56SPEAKER_02: So,
- 9:57SPEAKER_02: I mean,
- 9:57SPEAKER_02: the bottom line is these actors,
- 9:59SPEAKER_02: you know,
- 10:00SPEAKER_02: they appear to be doing this to just bring back money.
- 10:03SPEAKER_02: But obviously, you don't want the North Korean government in your system.
- 10:09SPEAKER_02: It's a terrible idea, right?
- 10:11SPEAKER_02: And we've seen some extortion scams from them as well.
- 10:13But the other thing I think that's really interesting is that
- 10:16it is just,
- 10:17it is everywhere.
- 10:18SPEAKER_02: I'm telling you,
- 10:18SPEAKER_02: we've seen situations where two,
- 10:20SPEAKER_02: these two have basically been competing,
- 10:22like two different North Korean IT workers have been competing
- 10:25for the same job.
- 10:26SPEAKER_02: I mean,
- 10:27SPEAKER_02: I think my,
- 10:28SPEAKER_02: one of my colleagues,
- 10:29SPEAKER_02: Charles Carmichael,
- 10:29will tell you that like nearly every CISO he talks to
- 10:34confidentially tells him that
- 10:35SPEAKER_02: they've had an incident.
- 10:36SPEAKER_02: It's unbelievable.
- 10:37And it's kind of hard to fight if you're a security
- 10:41organization because it's not
- 10:44SPEAKER_02: exactly a historically,
- 10:47SPEAKER_02: like it doesn't,
- 10:48SPEAKER_02: it's not exactly,
- 10:49SPEAKER_02: it's not similar to many of the sort of things that we fight every day,
- 10:52SPEAKER_02: right?
- 10:52You know,
- 10:53like zero day in the firewall,
- 10:54that's very cybersecurity focused,
- 10:56right?
- 10:56SPEAKER_02: And that's something that's well within CISO domain.
- 10:59But really the way you deal with this is really like it's an HR
- 11:03problem too,
- 11:03right?
- 11:04And so we hear about CISOs who have now reach out to the HR
- 11:07department and sort of
- 11:09SPEAKER_02: prod them towards shoring up their process,
- 11:12SPEAKER_02: right,
- 11:12SPEAKER_02: of hiring.
- 11:13SPEAKER_02: Right.
- 11:13SPEAKER_02: And that really comes down to doing background checking,
- 11:16SPEAKER_02: like physically checking identification,
- 11:18SPEAKER_02: right?
- 11:18One of the ways that they've gotten through in a lot of places
- 11:21is they're providing
- 11:22SPEAKER_02: like a picture of an ID and that picture could be edited.
- 11:25SPEAKER_02: It has been edited with AI.
- 11:27Bringing people into the office is another solution that we've
- 11:31seen,
- 11:31putting people on camera.
- 11:33SPEAKER_02: But a lot of that stuff has to be done in the HR.
- 11:36SPEAKER_01: Interesting.
- 11:37and i think it was just i was just blown away by kind of how
- 11:40almost simple it was
- 11:42for these actors to kind of get away with it and the way that
- 11:45they were kind of
- 11:46using people to have like up top farm yeah yeah so that's
- 11:50that's uh that's one of
- 11:51the really interesting sort of the soft underbelly like the
- 11:54area where we can
- 11:55SPEAKER_02: disrupt sort of nationally the best is these facilitators right and so
- 12:00SPEAKER_02: They are essentially going into,
- 12:01SPEAKER_02: you know,
- 12:02freelance sites in places like that and hiring people who are
- 12:06willing to host these
- 12:08SPEAKER_02: laptops that are coming from the company.
- 12:10And so the way that the scheme works is you are,
- 12:13North Korea might get the job.
- 12:15They might even use the facilitator to help them get the job by
- 12:17like getting going
- 12:18SPEAKER_02: on camera and things like that.
- 12:20SPEAKER_02: But ultimately,
- 12:21SPEAKER_02: like a laptop would be sent out and a laptop farm,
- 12:24SPEAKER_02: this person running a laptop farm might take receipt of the laptop,
- 12:27plug it in and then put a remote access tool so the adversary
- 12:31or the IT worker can
- 12:33SPEAKER_02: essentially come in from North Korea through this remote access tool,
- 12:36SPEAKER_02: do their job through their laptop and maybe the company's none the wiser.
- 12:40SPEAKER_01: Yeah.
- 12:42um what you know there's a lot of noise in this space right and
- 12:46we kind of the
- 12:47whole buzz this week has been agentic ai and those threats it's
- 12:50interesting that
- 12:50social engineering wasn't at the sort of the top one that
- 12:53you're concerned about
- 12:54what do you think we're not talking enough about in the
- 12:56industry or as a kind of
- 12:57SPEAKER_01: community that should be on the radar
- 12:59SPEAKER_02: You know,
- 12:59SPEAKER_02: I'll tell you,
- 12:59SPEAKER_02: I've had a lot of conversations with CISOs and security teams all week.
- 13:03SPEAKER_02: And this is literally like my last meeting of the week.
- 13:06SPEAKER_02: And so I'll tell you,
- 13:07the thing that I heard probably the most from security leaders
- 13:13is this insider
- 13:14SPEAKER_02: threat thing.
- 13:15SPEAKER_02: A lot of it's connected to this IT worker thing,
- 13:18SPEAKER_02: but organizations see that as sort of a larger,
- 13:21SPEAKER_02: a larger problem.
- 13:22And it sort of sits on the edge of this,
- 13:24like the cybersecurity space,
- 13:25right?
- 13:25SPEAKER_02: It's a cyber problem and it's kind of an HR problem too, right?
- 13:29SPEAKER_02: And,
- 13:30SPEAKER_02: you know,
- 13:30SPEAKER_02: I've been told,
- 13:30SPEAKER_02: you know,
- 13:31SPEAKER_02: a lot of organizations are,
- 13:32it's unfamiliar territory or they're not necessarily used to
- 13:36dealing with it.
- 13:37SPEAKER_02: But I think the IT worker problem has really shined its light on,
- 13:40SPEAKER_02: like,
- 13:40SPEAKER_02: this is something that we have to start thinking through.
- 13:42SPEAKER_01: Yeah.
- 13:43SPEAKER_01: What's kind of your top advice for security investments?
- 13:47SPEAKER_01: What's not talked about enough,
- 13:48but what are the kind of top things that they can do to kind of
- 13:50protect themselves?
- 13:51SPEAKER_02: Well,
- 13:51SPEAKER_02: you know,
- 13:52SPEAKER_02: if you're,
- 13:52SPEAKER_02: if you are doing this at all,
- 13:55SPEAKER_02: right,
- 13:55SPEAKER_02: I think that there's two ways that,
- 13:57SPEAKER_02: two important things to think about.
- 13:59SPEAKER_02: One is sort of brilliance of the basics, right?
- 14:01SPEAKER_02: You've got to, you've got to cover the fundamentals.
- 14:04SPEAKER_02: But you really also have to start with what's your threat model, right?
- 14:08If you are involved in the defense industrial base,
- 14:11you have a serious nation state threat.
- 14:14SPEAKER_02: If you are a,
- 14:15SPEAKER_02: I don't know,
- 14:16a medium-sized parts manufacturer or something like that that
- 14:20has no defense sort
- 14:21SPEAKER_02: of relationships,
- 14:23SPEAKER_02: your major concern is probably criminal,
- 14:26SPEAKER_02: right?
- 14:26SPEAKER_02: Yeah.
- 14:27SPEAKER_02: You know, I talked about sort of that evolution of cybercrime.
- 14:30SPEAKER_02: One of the biggest evolution,
- 14:31SPEAKER_02: you know,
- 14:32or one of the biggest changes when actors stopped sort of
- 14:35focusing on fraud and
- 14:36then moved towards extortion and ransom is that they started
- 14:38targeting everyone,
- 14:39SPEAKER_02: right?
- 14:40SPEAKER_02: And so...
- 14:42based on what the adversary is doing and what is the most
- 14:45critical to your business.
- 14:46SPEAKER_02: And that might be the same for everybody,
- 14:48SPEAKER_02: but like in most organizations,
- 14:49it's going to be the threat of disruption or the threat of your
- 14:52information being
- 14:53SPEAKER_02: leaked on the open internet,
- 14:54SPEAKER_02: right?
- 14:55SPEAKER_02: Which would be usually a serious risk to your brand.
- 14:58SPEAKER_02: And the defense against that is backups,
- 15:01SPEAKER_02: preparation,
- 15:02SPEAKER_02: you know,
- 15:02SPEAKER_02: and mostly,
- 15:03SPEAKER_02: you know,
- 15:03SPEAKER_02: like the basics.
- 15:05SPEAKER_01: Yeah.
- 15:05SPEAKER_02: So like MFA and...
- 15:10SPEAKER_01: What kind of excites you most about the cybersecurity space today?
- 15:14SPEAKER_02: Well, I think the AI stuff is really interesting.
- 15:18SPEAKER_02: We've clearly crossed some river.
- 15:20SPEAKER_02: And we're crossing it so fast, it's crazy.
- 15:22SPEAKER_02: But we've crossed something right here.
- 15:24SPEAKER_02: And there's a lot of hype, right?
- 15:29SPEAKER_02: And it's hard to sort of peel through that.
- 15:32SPEAKER_02: And I got to tell you, I've heard use cases that I think are ridiculous.
- 15:36SPEAKER_02: And I've seen stuff that is mind-blowing right at the same time.
- 15:40SPEAKER_02: And it's going to take some time for the reality to shake out, right?
- 15:47SPEAKER_02: But I think we're going to get there.
- 15:48SPEAKER_02: I do think that... Two things.
- 15:50SPEAKER_02: I think that the adversary is going to figure out new ways to use AI,
- 15:54SPEAKER_02: and it's going to help them scale their operations,
- 15:56SPEAKER_02: and it's going to make them better.
- 15:57SPEAKER_02: And I think that we are...
- 15:59SPEAKER_02: whether we like it or not now in an arms race with them.
- 16:02SPEAKER_02: And so we're going to have to get better.
- 16:03SPEAKER_02: We've always, by the way, needed to get better.
- 16:05SPEAKER_02: It's not been easy.
- 16:06SPEAKER_02: I think the adversaries always had sort of an advantage over us, right?
- 16:09Like,
- 16:10you know,
- 16:10I think most people,
- 16:12the defender's dilemma,
- 16:12most people would agree with that.
- 16:14SPEAKER_02: But also AI could be the solution that we've been looking for.
- 16:18controlled it's control essentially controlled by by
- 16:21organizations outside of the
- 16:23the you know adversary space um we control a lot of the data
- 16:27that could make it
- 16:28really good right and really effective and i think that uh you
- 16:32know as the use
- 16:33cases are coming and we were already seeing some really cool
- 16:36ones i think like the
- 16:37one still the one i point to the most is like it's very very
- 16:40helpful with reverse
- 16:41SPEAKER_02: engineering
- 16:42SPEAKER_02: And I used to run a cyber esponage team.
- 16:44SPEAKER_02: We just focused on cyber esponage threats.
- 16:46And usually the work or the sort of the workflow was sometimes
- 16:50some kind of weird
- 16:51malware came in and you'd have somebody spend two days trying
- 16:54to figure out what
- 16:55SPEAKER_02: the heck that malware did.
- 16:56SPEAKER_02: And after that, you would determine whether or not it was interesting.
- 16:59SPEAKER_02: Anyway,
- 16:59we've got more and more ability to sort of break that open and
- 17:02figure out whether
- 17:03SPEAKER_02: or not it's interesting a lot faster.
- 17:04SPEAKER_02: Right.
- 17:05SPEAKER_02: And that's really cool.
- 17:06SPEAKER_01: Yeah,
- 17:06SPEAKER_01: I've heard it said that,
- 17:07SPEAKER_01: you know,
- 17:08so far the defenders have had the advantage with AI over the
- 17:11adversaries and
- 17:11SPEAKER_01: they've got,
- 17:12SPEAKER_01: well,
- 17:13SPEAKER_01: for Google,
- 17:13SPEAKER_01: the data,
- 17:14SPEAKER_01: you know,
- 17:15SPEAKER_01: that is the advantage with AI,
- 17:16SPEAKER_01: isn't it?
- 17:16SPEAKER_01: It's only as good as the data you've got.
- 17:18SPEAKER_02: Yeah, and that's a really, really important piece.
- 17:21I think,
- 17:21like,
- 17:21you know,
- 17:23the technology obviously is super important,
- 17:25right?
- 17:25SPEAKER_02: But it's just not, the technology is only as good as the data.
- 17:28SPEAKER_02: And I'm also concerned that,
- 17:31SPEAKER_02: I think the adversary is going to learn that about the whole AI problem,
- 17:34SPEAKER_02: right?
- 17:34SPEAKER_02: And so one of my concerns is that,
- 17:36SPEAKER_02: Ultimately,
- 17:37SPEAKER_02: as more adversaries and others get access to AI,
- 17:45the realization will be that the data is increasingly a massive
- 17:49differentiator and
- 17:49SPEAKER_02: it will actually create an impetus for more cyber espionage or theft.
- 17:56SPEAKER_01: And just,
- 17:56SPEAKER_01: yeah,
- 17:57SPEAKER_01: final question,
- 17:58SPEAKER_01: I guess,
- 17:58SPEAKER_01: just looking ahead,
- 17:59what do you kind of predict is going to be the kind of the way
- 18:02the threat landscape
- 18:03SPEAKER_01: looks?
- 18:03SPEAKER_01: You know, when we're here at RSA next year, do you sort of see any more?
- 18:06SPEAKER_02: You know,
- 18:07SPEAKER_02: I'll tell you,
- 18:07SPEAKER_02: you know,
- 18:08every year you get asked that and it doesn't necessarily change
- 18:12super rapidly,
- 18:14SPEAKER_02: right?
- 18:14SPEAKER_02: Like we were talking about the network edge a few years ago.
- 18:17yeah right and it's still still in play what changed though was
- 18:21like we could see
- 18:22these like the the industrialization of the play there right
- 18:26and so like almost
- 18:27like what changes the momentum against the edge right and so
- 18:31i'm hoping that we've
- 18:32already named something that that you know like that is a
- 18:35problem but maybe you
- 18:36know we were on the early edges of it right and we're seeing
- 18:39those things instead
- 18:40SPEAKER_02: of catching us unawares yeah john thank you so much thank you
- 18:46Bye.