00:00I know you've had a couple of partnerships already announced, including with Amazon,
00:03as well as a few other companies that kind of predated the release of Mythos. But I do want
00:08to get your thoughts here on the capabilities, at least as far as what you know them to be,
00:13and what type of threat, if at all, it does actually pose.
00:18So these new models, whether it's Mythos model or the model from OpenAI,
00:22they are powerful to find new software vulnerabilities. But the issue is not lack
00:28of finding software vulnerabilities. In fact, every large enterprise has so many software
00:34vulnerabilities, they can't patch them, they can't fix them. Now the number gets even bigger.
00:39That means they need to do something to take care of them. This is where Zscaler comes in.
00:45Our zero trust security can actually hide your applications from the bad guys.
00:50So if they can't reach you, they can't breach you. So companies like Zscaler are needed far more
00:56in this world than they were before. Do companies understand that? I mean,
01:00potential clients understand that? Because there are a lot of people that would say that even with
01:04the threat of some of these more powerful models, the idea is that maybe you don't necessarily have
01:10to rely on the traditional cybersecurity companies. You can almost, for lack of a better phrase,
01:15fight an LLM with your own LLM.
01:18I think the world is simplifying it. Investors are not getting it. Enterprises are getting it. I have
01:26had two dozen calls with the CIOs and CISOs of large enterprises in the past two weeks. They're all
01:33asking for help. They're saying, how do I handle this stuff? LLMs can help you find vulnerabilities.
01:40They can help attackers to figure out how to exploit them. And they are going to fix these vulnerabilities
01:47on their own. Help is needed. We had a webinar yesterday with 2,000 enterprise CISOs and CIOs
01:55attending. And the main topic was, how can Zscaler help? And the help is in many ways. There's no one
02:02single silver bullet hiding their applications behind Zscaler. Exchange is one. Prioritizing the risk of
02:10those vulnerabilities is two. Implementing zero trust segmentation is three. So multiple things
02:16need to be done. We are engaged with our customers to help. Well, something I'm curious about, Jay,
02:21is, you know, you mentioned when you're talking to these, you know, chief security officers,
02:26chief information officers, what sense do you have of, you know, whether they're partnering with
02:31multiple cybersecurity firms or, you know, exclusively with Zscaler? Because you could see a case for,
02:38you know, just given how big and expansive the threat landscape is, maybe you want to have
02:44multiple points of coverage. Yes. In general, there are multiple partners, but it's not 20.
02:52Okay. Generally, they look at a handful of partners. There's a certain part of cybersecurity.
02:58We do extremely well. And there are complementary other vendors who do certain areas. The goal is
03:04bring together the talent and expertise to make sure enterprises safe. The bad guys can fail 99 times.
03:12They only need to succeed once. We or enterprises can succeed 99 times and fail only once. That's a
03:18problem. That's why enterprises are taking pretty seriously. CEOs and board members are taking seriously.
03:26So there's never been bigger need and demand for services like Zscaler than it has ever been.
03:32Yeah. And of course, it's that one time that is what's the broader public is made aware of. I am
03:39curious.
03:39I mean, when it comes to a mythos, when it comes to some of these other models that we're seeing
03:42come out,
03:43these advanced models, I mean, what does it look like on your end, how you evaluate them, what your due
03:50diligence
03:50process looks like?
03:51So we have been working with Entropic and OpenAI, both of these companies, early stage, before release, so we can
04:00identify some of these vulnerabilities and fix them in our platform as well, because we are protecting over 45%
04:08of Fortune 500 companies.
04:10And we're making good progress. But the next step for us is to take the same kind of protection and
04:16help to our customers. And that's what we're doing. Now, these models will become generally available. Today, we are talking
04:24about Entropic and OpenAI, which are both our partners. Tomorrow, Chinese companies will come up with some of the same
04:31models.
04:31So we need to be ready for it for enterprises. It is definitely an area where it's a much heightened
04:40risk, and we need to work together to minimize it.
Comments