Inside the Enterprise Browser Rebuilding Security for the AI Era | Bradon Rogers, Island

AI is moving faster than enterprise security systems were designed to handle. In this episode of Eye on A. I., Craig Smith speaks with Bradon Rogers, Chief Customer Officer at Island, Island about how companies are struggling to govern the rise of AI agents, browser-based workflows, and unsanctioned AI tools inside the workplace. The conversation explores why traditional "block-and-control" security models are breaking down and how a new approach, embedding policy directly into the browser and user workflows, may offer a path forward. It also dives into emerging risks like prompt injection and autonomous agent behavior, and why enterprises are increasingly becoming multi-AI environments by default. Subscribe to Eye on A. I. for weekly conversations with the people building and deploying the future of AI.

Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.

• 0:01Craig S. Smith?

  Can you talk about the risks that are developing with AI browsers and autonomous agents as they are designed by consumers or by users, but then are being integrated into the enterprise? There are a lot of agentic tools that are not being developed inside the enterprise.

• 0:21Speaker 01

  The agent is working over an application. There's hidden instruction buried in the application. The agent goes and follows those instructions, thinking those are the human giving them instructions in the process. The agent's nothing more than me hopping out of this seat and you see an empty chair, but the agent's doing work on my behalf. And we want to have policies that are built and let the agents operate in the confines of a given policy, no more, no less, and make them live in a place where they can be enterprise ready, no matter what the AI provider is that you're using.

• 0:44Craig S. Smith?

  Automation enables teams to build and run on-demand agents that operate at speed across enterprise applications. How does Island keep track of what's being built by teams in an enterprise? I usually start by having you introduce yourself to listeners and give as much of your background as is relevant and how you got to Island, and then what Island does, and then we'll start talking about keeping the agentic world safe.

• 1:28Craig S. Smith?

  I mean, that seems to be one of the focuses.

• 1:32Speaker 01

  Yeah, thanks, Greg. I'm Braden Rogers, Chief Customer Officer with Island. The role that I play is anything that's technology and engineering-centric that's in the field that touches customers is the world I work within every single day.

• 1:45Craig S. Smith?

  Island just announced some new functionality. Can you talk about that?

• 1:52Speaker 01

  Yeah, 1,000%. First of all, it's important to acknowledge the incredible innovation that we're seeing every single day and the breakneck speed with which everyone's experiencing AI. And you know, it's everything from people leveraging generative prompts to exchange information and to make their work more effective. And it's obviously led over into some really interesting areas around automation and agentic workflows as people have engaged it. And you know, I think it's important to acknowledge that many of the players that are doing this work are doing some incredible work. A lot of your different providers, the big brand names that everybody

• 2:36Speaker 01

  knows well that are the AI players. But a lot of the work that we see from that ecosystem is very consumer driven. There's nothing wrong with that. At the end of the day, they're chasing billions of users around the world. And you know, but at the end of the day, when you're a large financial services firm, or a large healthcare provider, you certainly have concerns with these things, you know, making their way into your environment. And then also, the other thing is, you want users to be empowered. You want to wrestle with this consumer need, this consumer thing. And there's some goodness, really, there's amazing goodness with it. And you know, we certainly believe AI is here to kind of help us in the future, not to decimate

• 3:21Speaker 01

  our futures. And but we want to be able to harness that and leverage that power and enable our users very effectively. And for us, you know, there's an avenue for, it's really focused around enabling the user, but most importantly is allowing cybersecurity have a role in one of those rare times in the universe of cybersecurity of being a true enabler. Let's weave AI into the fabric of the user's workflows, and let's do it safely. And let a cybersecurity team play a very deep role in that enablement. And that's what we announced yesterday with a series of capabilities I focused around the universe of AI that's square in the laps of everybody in the enterprise today.

• 4:05Craig S. Smith?

  Yeah, can you talk about the the the risks that are developing with

• 4:13Craig S. Smith?

  AI browsers and autonomous agents as they are are designed by consumers or by users, but then are being integrated into the enterprise? Is is that the issue really that there there are a lot of agentic tools that that are not being developed inside the enterprise but are being brought into the enterprise? Or or is it even for for tools that the enterprise is building inside its its the company?

• 4:50Speaker 01

  Great question, Craig. I I think it's important to I'm I'm gonna go up a level real quick before we go into the Gentic side, because I think it is it's this series of AI resources. Sometimes there are Gentic resources, sometimes there are environments where users are engaging our corporate data into a prompt. Sometimes there are developmental tools with, you know, various forms of vibe coding resources that help us speed development of applications. But there's a whole series of resources people are bringing to the table for their jobs. And, you know, I think it's I think it's fair to say that, you know, your average user can find immense value in AI as a as a bit of a cheat code for the things that were difficult

• 5:31Speaker 01

  in their job, you know, and, you know, sometimes it's about the knowledge that it can bring forth that otherwise they they may not have expertise in. Sometimes it's about automating their workflows. Sometimes it's building an app that, you know, now now a non-technical human being can create an application that's very sophisticated. But these things are coming at the organization so fast and so furiously that the organization is is is faced with, you know, you just because these things have a lot of promise did not mean you abandoned your your fundamental requirements. If you're a financial services firm or healthcare provider, any other technology or enterprise, I should say, didn't mean you abandoned your requirements. And by the way, they're more important than ever, your regulatory bindings that you're having to live by.

• 6:13Speaker 01

  Yeah, it's even harder now to live by those bindings with with this universe of chaos coming to the table. You you your data protection needs that you had. You want to you want your users engaging in the sanctioned stuff, not this shadowy universe of stuff. And so at the end of the day, you know, it's it's the chaos, the orgs are trying to trying to wrestle with the chaos. But then while they're wrestling with the chaos, not turning into not turn into the no police. Because, you know, I think it's a very fair statement that that a lot of the world wants to be AI empowered. Think of a lot of a lot of organizations are thinking about how do we be forward leaning? How do we think AI first? And thinking AI first is not saying no to everything for your end users.

• 6:55Speaker 01

  It is shepherding and weaving in the sanctioned world of AI into their workflows and empowering them to use AI as their first first thought process. When they go to to to get into something that's creative or building something or being more faster at their job. And so for for my universe, or at least the way I think of it is, I want to think about, you know, AI empowerment first and AI mindset first. For people that are thinking that way. And how do I encourage a user who otherwise might not think that way to naturally gravitate that way? Because it's now in their natural workflows to make their job easier, but it's also safe. And that is, there's a number of different angles in this whole thing.

• 7:33Speaker 01

  There is definitely the chaos that is the onslaught of stuff that's coming at the organization that that it's hitting them before they've gotten their sanctioned answers and their their core strategy in many cases.

• 7:42Craig S. Smith?

  Yeah, explain explain on how how island works with the enterprise.

• 7:48Speaker 01

  Well, I I think it's important to highlight what we built at Island. Part of AI, just, you know, if you think about the the workspace for the user, the majority of the work of a user is spent day in, day out, using a browser. And it's a great piece of technology. It's one of the most important pieces of tech that has been in been around since the advent of the Internet, obviously. And at the end of the day, the evolution of the browser, it we for our original use was consuming information. And then all of a sudden, it became an application delivery app. We receive our applications in that window. Yet the mechanics of that browser are not built for that purpose. They're actually built for shopping. I mean, you think about the targeted advertising and the focus around the analytics around the end user. It's a consumer focused need and that's okay.

• 8:34Speaker 01

  It's just that we, we, we said, look, there's some potential kinetic potential, you know, pin up in that browser. What if we could uncork that potential and transform the browser actually, actually into an application delivery platform? We're building mechanics and policy to live locally for the experience of the end user. Things like, you know, building in data protection, building in contextual awareness, building in protection for the user as they, you know, engage the broader, uh, you know, internet. But we want to use policy locally to govern that experience and then also make the user more productive, productive in a basic way, just as a starting point. They already know how to use a web browser. We don't have to train them. So what if you could just start putting mechanics of their work right at their fingertips in the, you know, again, woven into the fabric of the browser? And that's where we started the journey.

• 9:18Speaker 01

  And then, you know, along comes AI and its natural habitat as a browser, you know, for much of what the user engages, AI's universe is browser centric. So as a starting point, you know, then starting for us is weaving the AI mechanics into that, that user's workspace. So the user has seamless access, but always bound by the policies that we had in play before, the data protection elements. Shepherding them to the right tenants, you know, the, the right corporate co-pilot versus personal or the right, you know, uh, Gemini environment versus personal. Or better yet, let them go to personal, but don't let the corporate data spill over the boundary, you know, into the tabs that are the personal tenants. So all of those are, you know, kind of fundamental mechanics.

• 10:00Speaker 01

  That is what we launched yesterday is about the empowerment of the end user to take advantage and leverage AI. And then taking it a step further is taking it into areas for agentic workflows, allowing agents to have the ability to operate in the workspace of the user, but have appropriate guardrails. Uh, these are all very important foundational things and you, you certainly seen, um, you know, stories. It didn't take a lot of search through Reddit threads and things where people have experimented with these agentic browsers. You see, you know, analyst firms like Gartner Publishing Research saying, hey, like be careful. Stay away from these agentic browsers in the enterprise for now. Um, and, uh, you know, there's, there's a reason for that. At the end of the day, we gotta get, we gotta put control around it and, you know, minimize the chaos, but also again, back to empowering users.

• 10:44Craig S. Smith?

  Well, let's talk about some of the new, new products. AI protect, which allows safe usage of any AI application, consumer or enterprise. How does that work? I mean, where is that? A plugin to an AI browser or is it something more horizontal?

• 11:05Speaker 01

  Yeah. So if you think about how the user may engage AI, they're going to engage AI in their consumer browser, their Chrome or Edge. They're going to engage AI in an enterprise browser like Island. Are they going to use localized applications such as, you know, their local clients for chat GPT as an example, or whatever the tooling may be? For us, what we announce is the ability with AI protection is to live anywhere the user lives. Whether they're working in the enterprise browser. If they're working in a consumer browser, using the island extension to assert our policies inside the consumer browser. And then using Island Desktop, which is a service that lives locally on the host to assert policy outside of the browser and govern the usage of AI resources.

• 11:48Speaker 01

  No, again, no matter where the user's working, govern those AI resources accordingly. Uh, that's exactly what AI protect is. AI protect is really two fundamental, very foundational things for Ireland, but very difficult for most status quo traditional providers. Um, it is, uh, it is about having visibility, understanding who's using what and building policies. And many of those policies, while they may seem, uh, you know, very simple, like many upstream technologies have a hard time saying, yeah, let's let users use personal Gemini. And the company data is not going to spill. There's no, no risk of company days. Though what they do is they, the strategy, what you see oftentimes is a strategy, um, that it's, it's like, uh, you think about the pinch points.

• 12:30Speaker 01

  The upstream pinch points, we're passing traffic back through a cloud and inspecting at a pinch point. Their their strategy has to be a block page. They got to block the personal stuff. That's assuming I can even identify the tenancy accordingly. So so that but we take a quite a different approach to empower the end user. If they're using their consumer browser, they're using their enterprise browser, using the local clients, let them use personal stuff. But recognize when they're using the corporate resources and bring them into the fold into the boundary. So they can have freedom of move it. And the agents fall into the same types of flows as well. Again, the agent's nothing more than me hopping out of this seat and you see an empty chair, but the agent's doing work on my behalf.

• 13:09Speaker 01

  And we want to have policies that are built and let the agents operate in the confines of a given policy, no more, no less. But again, make them, you know, live in a place where they can be enterprise ready, no matter what the AI provider is that you're using.

• 13:22Craig S. Smith?

  So for example, Slack is becoming agentic. How would Ireland handle that? I guess Slack is delivered. Either through the browser or through an app. This desktop of functionality, how does Ireland like wrap around that that sort of thing?

• 13:43Speaker 01

  You know, an app like Slack or any other app, it's going to live in the browser. It's going to live as a local thick app. And the user engages those interfaces. And those interfaces could, they may be making web calls over here. The thick clients can be making web calls too, because a lot of times thick clients are kind of browsers in disguise as electron apps. But then many times in the universe of of AI, you also get into conversations that are, you know, other protocols, more modern AI centered protocols like MCP, you know, calls. And so, you know, for us, the the ability for the user to use whatever client they're using, whether it's in the browser or a thick client. And asserting our policy over those things, including MCP calls at the end of the day.

• 14:27Speaker 01

  So that when the agent is engaging something, the agent's even doing it within the permissibility scope of what it should should be doing. Not falling victim to, you know, I should go over here now and go to personal Gmail and start grabbing tokens from, you know, out of here or taking corporate tokens from this environment over here. And, you know, those types of things. That we've seen again that have been proof of concepts over and over again, shown around a lot of that agentic world, you know, going amok. And again, it's early. So you know, a lot of that stuff is naturally going to be experimented with. People are going to push the boundaries. We just want to make sure that the enterprise can take advantage of that stuff, you know, today. And and feel safety in doing that as well. So the Slack clients, agentic workflows, just like any other agentic workflows, whether it be

• 15:09Speaker 01

  a browser based agentic workflow or something outside the browser is really the scope of what you know, we're focused on tackling.

• 15:15Craig S. Smith?

  So the listeners understand the risks that you're addressing. What? Give some examples of what can go wrong.

• 15:24Speaker 01

  First of all, I think it's important to point out that, um. You know, going back to the conversation of that these, a lot of these things that's focused around the consumer need. Um, so perfect example, you've seen examples of, uh, you know, there's in search of a couple of red threads, you'll see examples of of people exploiting the agentic workflows for prompt injections, for example. Um, the agent is working over an application. There's hidden instruction buried in the application. The agent goes and follows those instructions, thinking those are the human giving them instructions in the process. Uh, and then the next thing you know, the agent's over here grabbing, you know, tokens, you know, corporate tokens out of a Gmail account that were sent via email and it shouldn't be doing that. But at the end of the day, it's doing it and it's following what it thinks are instructions.

• 16:07Speaker 01

  Um, we've got users that are very basic level type stuff. Users taking corporate data, launching a personal app, AI app, taking corporate data over to that personal AI app. In a place we don't really want it because we just lost custody of our data because now it's now it's gone into that AI providers ecosystem. Um, it it comes from a lot of different extension ecosystem. You know, you've there's, I think I read the other day, there's like 18,000 plus classified extensions that are classified as AI. That's a shadowy worm of things that wind up in your environment. You don't even realize they're AI living in your environment. There's another angle we've got to think about how we govern. So there's, there's all kinds of risks of, again, it's, it's a lot of it's centric around the data that you have in your

• 16:50Speaker 01

  organization, but in the agentic world is also centered around what the agent may do autonomously on your behalf, thinking it's doing the right work and inadvertently doing something it shouldn't be doing. So we want to build guardrails and policies around those agents, just as we've have around the end users. Again, a very smooth experience, but we want to let the user also, by the way, because the agent doesn't exist independent of a user. A user sends instructions. We'll let agentic stuff be presented to the user to be able to take advantage of these things and not have to worry about the risks themselves either.

• 17:19Craig S. Smith?

  How, how does it work under the hood? Because, yeah.

• 17:26Speaker 01

  Yeah, 1,000%. So anywhere you assert an island policy, whether in the browser, or whether you assert it inside of the extension or inside of Island Desktop.

• 17:37Speaker 01

  Again, you think about the movement of, of an agent. The movement of the agent is moving in a very similar way akin to the end user. I'll use an example. Yesterday I was doing some agentic work and I asked it to go to a social media platform and find and connect me to certain people that were in the sphere of a particular topical area that I wanted to learn about. And I let the agent go and do that work. At the end of the day, the agent would be bound by a policy that lives in the browser and that give an example. So if the agent decided, oh, it saw some instructions to go over and take this data from here and go over to the boundary to the personal stuff or some other area it shouldn't post the data to, the policy lives in play and keeps it from doing that. So the agent would still bump into the bound, the guardrail that was there for the user. And maybe we may even have more

• 18:19Speaker 01

  rigid guardrails for the agent for certain things as well. So the, the guardrails could, could confine the agent to specific sets of, of work, like, you know, only automate this one thing, but don't go out beyond the confines of that. So policy lives locally. That's really important. Lives locally in the browser for those things, lives locally in the extension, asserting itself in the consumer browser and lives locally on the host with Island Desktop. Again, those were all announced yesterday as well, part of the Island AirPods platform. And then so as a result, when the agent It starts asserting its will. Oh, let me go touch this thick application outside the browser or touch this folder in the file system. The Island desktop is asserting its policy for the agent, just like it would a user in that particular case.

• 19:02Speaker 01

  And then when the client is engaging some MCP resources as well, we assert our will in that process as well. So that again, anything going into the MCP flow or coming out has been governed or at least audited at the appropriate level for those engagements.

• 19:16Craig S. Smith?

  Audited and and logged and so that so that there's a virtual paper trail.

• 19:23Speaker 01

  That's right.

• 19:23Craig S. Smith?

  Of everything that goes on. How is it configured? I mean, do does the enterprise? Yeah. How do you work with an enterprise?

• 19:31Speaker 01

  Yeah.

• 19:31Craig S. Smith?

  That's, yeah.

• 19:33Speaker 01

  Well, I think that the one thing that's important to point out about what we've built for the past five years. It's a very novel approach to asserting our will, our organization of will. By the way, the will is not just cybersecurity. It's our will for productivity resources for the end user and automation resources for the end user and digital experience and stuff like that. But it is a very novel approach. You know, you think about it, you already know how to use a browser. Today, Craig, if I had to onboard you in the universe of Ireland, I would tell you to go to the App Store of your device of choosing, or go to the Island download page and download Island. And something tells me somewhere in your past, you've actually downloaded and installed a browser, probably many times in your life.

• 20:16Speaker 01

  You launch Island, you log in, the org has built a policy that the moment you log in, the browser and our services instantly reach up to the cloud, grab the policy that the organization built. And boom, all of a sudden, your application entitlements are provisioned for you. The policies provisioned locally, the the automation resources, et cetera, are all provisioned locally. So for you to onboard yourself as an end user, it's a self provisioning exercise because you need to know two things in the universe of Ireland as an end user, especially if you're self onboarding. You need to know how to go to an app store and download an app, in this case, our browser. Second thing you know, and you know how to do is log into the single sign on provider for the org that you work for. If you've ever logged into your Entra ID or Okta environment, you're done.

• 20:57Speaker 01

  All of a sudden, context based on the positioning of the policy that the org built is now staged for the end user. So the end user's experience is literally nothing more than engaging the familiarity of a browser. And this is why we don't require any training for the end users either. It's they just now see it and they go straight to work instantly.

• 21:14Craig S. Smith?

  Wow. And so Ireland can can find and read all the the policy documents required to to govern agentic uses and that sort of thing.

• 21:28Speaker 01

  Yeah, so you obviously the org is going to build a policy ahead of time. The policy is going to have all this. And by the way, many of our Ireland customers already have these data protection policies. They've been building them, working with Ireland for years and deployed at massive scale. So they built these things. And so now they've already built them. Now AI comes into the fold. You simply empower, enable the capabilities to integrate to your proper AI providers. Now, AI is the AI is elements are living, slipstreamed right into the existing policies that the user already has. And as we continue to evolve capabilities, they'll always just evolve and and begin to slipstream straight into the capabilities of the end user. And that includes even new apps in the AI universe. You know, one of the things we launched, announced yesterday was this area of around AI publish.

• 22:12Speaker 01

  You know, you want to be able to leverage, you know, resources like vibe coding tools like lovable and cursor and, you know, Claude code, stuff like that. You want to build those apps and be able to deliver them and actually have them run and then have those apps slipped right into the island policy as people deploy the apps and deploy them at the same speed that they're deploying them in the vibe coding tools that they're creating. So the speed is matched for deployment. That is happening within the the delivery, but cyber cybersecurity policies are never abandoned in that process because it's literally slid right into the exact policies you already have. Because again, that world is very much, even though they're AI coded apps, they're web based apps. They're they're they look and feel web friendly. You can build thick apps, but you know, most people are building web centric type apps. They're very rich type stuff.

• 22:57Speaker 01

  And again, you don't have to walk away from all the stuff you built in the past. It was important for you.

• 23:03Craig S. Smith?

  Yeah. And but this is a B to B product. It's it's not B to C. So or or can it be B to C? I mean, you you were saying that you go to the App Store and and download the app. How do enterprises use it? Do they

• 23:25Craig S. Smith?

  have a is it buy seats or? Yeah.

• 23:31Speaker 01

  Yeah, it's a, it is definitely an enterprise focused. And I say enterprise, it's not just for the large end of town. It's for all end of town too. So we've got orgs that are, you know, small orgs that are getting value out of Ireland. Cause you know, that small org, like let's say you're a small bank with 200 employees. You have the same regulatory bindings that a large massive, you know, bank has and has to live up all the federal government requirements and then maybe international requirements as well. Um, you still have that, but you can't afford the same stack they have. You don't have the same practitioners and resources. So you need to do more with less. And we empower that for the smaller org. So it's, it's, it's, it's a, it's an organizationally focused effort, not focused on consumers. Um, it's focused is around, you know, user based pricing at the end of the day. So it's pretty simple.

• 24:17Speaker 01

  You know, at the end of the day, a, uh, a, um, a user can use as many apps as they want. They can use as many browsers, as many devices. That's not a consideration. It's not a concern. Um, so we just make it really simple for people to be able to consume at the end of the day.

• 24:31Craig S. Smith?

  But let's say an organization with 200 people. Uh, do you charge by seat and then, uh, you deliver an app for each 200 or each of the 200 people? Or they, each of the 200 people downloads the app on, uh, on, uh, island, on the, on their enterprises license. How does the organization?

• 24:57Craig S. Smith?

  Get island into its employees' hands?

• 25:01Speaker 01

  Great question. So there, it goes one of two ways. I use the extreme example earlier when I thought about Craig going to the app store and downloading the app. That, that's often what we see in the unmanaged real estate, a third party contractor. Um, you know, unmanaged device, a third party BPO, business process outsourcer. You outsource your call center, a third party entity. Um, you could, because you don't have the ability as, as an organization to force install software on somebody's machine you don't own. That breaks the laws of physics. So, so we give an easy way to onboard for the unmanaged. But on the managed side, you deploy like you do any other software, leveraging your existing, you know, software deployment methodologies that have been in play for years. So we just slip

• 25:43Speaker 01

  Island into those processes. And oftentimes, in many of the works where Wiley deploy, we're just part of their core build when they ship a device to somebody. Island is just part of the foundational build for those people in that case. And then to onboard themselves, they just go log into the SSO and they go to work.

• 25:56Craig S. Smith?

  Automation enables teams to build and run on-demand agents that operate at speed across enterprise applications. Does, how does Island keep track of what's, what's being built

• 26:18Craig S. Smith?

  by teams in an enterprise? Or is it, as long as you're working on an Island desktop, or, or in an Island browser, everything is automatically covered? Or, or does somebody have to submit a new agentic capability to, to Island to be reviewed?

• 26:47Speaker 01

  There's, there's a couple of ways to go about that. The, the, the most, the most effective way is to leverage AI automation in the construct of provisioning workflows for your end users that are, that are enterprise-grade workflows. They're, you know, the thing you want is you want predictability out of your workflows for your end users. Like, you want to take, let's say, a call center worker, and you know that it takes them 10 minutes to accomplish this task when they, when someone makes a call in for a reservation, or for getting an account status. You don't want them to have to go through and learn that process. And sometimes those processes could be across three or four different apps.

• 27:30Speaker 01

  They got to learn, you hire the new, there's the onboarding experience that they got to learn the process. And especially if you're out, if you're outsourcing to a third party, it's even more painful. It wouldn't be great if they could just push a button and it goes and accomplishes those tasks. So a big part of what we do is, is pre-staging the workflows that we publish. Publish those workflows to the audiences of users. And then when the user visits that part of the application, serve up reminders based on the, based on the, uh, the learnings we have about that user. The, like for example, you know, one of the things we're really heavily focused on is learning the role of the end user, uh, understanding the nature of the end user. But use a parallel.

• 28:12Speaker 01

  You know, when Elon Musk, you know, has gone down the path of, of building autonomous vehicles. He didn't put, you know, um, inspection points uh, at every intersection and watch how cars drive and drove at every intersection. That's how the existing world of technologies, you know, when you, especially when you think about the cyber universe would approach the problem. They would, you know, they get the pinch points, the upstream cloud proxies and, you know, just everything's being zeroed in on a specific pinch point so it can explode outward. You're, it'd be, it'd be akin to Elon Musk trying to learn how people drive cars by putting cameras at every intersection. Didn't do that. They sat in the seat of the car. They observed how traffic works as you drive down a street because not every street is an intersection

• 28:57Speaker 01

  where a camera would be in the, in the pitch point example. And so, but what you do is you start observing how the, what traffic looks like, what patterns look like, how, you know, and then you, you capture that data for that learning. And then you, you, you instrument that in a way where you can now automate vehicles. There's a lot more that goes into that, obviously. We're doing the same thing. We want to do the same thing for the end users. As the end user engages apps, observe their workflows, understand what they do, because that's not a packet level engagement living up in the network somewhere. Uh, if I observe their workflows, then what I can do is learn the nature of that user, the type of work they engage. Oh, that's a nurse practitioner and we see them doing this, this and this. Well, wouldn't it be make sense for us to present to them, um, proactively on the screen?

• 29:41Speaker 01

  Here's some prompts you should use to make your job easier. Oh, by the way, here's some workflows that we provision for the, for the organization provision for you. So you don't have to learn the check-in process for this patient when they enter the doors of the facility. You push these buttons and it goes through the five systems to do the, the basic work, shrinks tasks down from. Five minutes of work to thirty seconds or ten seconds. And so we get we get predictability out of those workflows that that are reliable. It don't go off and go awry and start doing other stuff. And we leverage the agent in the process to handle the delicate parts. Where you know, things in the application may have changed. Someone changed the the name of a button on a screen. Your traditional automation technologies would have broken with that stuff. Leverage agent for those types of delicate tasks.

• 30:25Speaker 01

  But that way, the agent doesn't go amuck. It stays in the cons, the course of the of the specific task is engaged in. And then also, one of the most important things is on the back end, measure that. Because it's important to understand, are we getting benefits from these things? And if you can turn around and say, yeah, three thousand call center workers. Use these ten workflows repeatedly on average of thirty, forty times a day per worker. There's a lot of time savings that come along with that stuff. If you can on the back end, give the give the value of that. And that's a heavy, heavy focus is not just providing agentic workflows. But making them enterprise grade, making the enterprise ready for the right audiences of users. Based on the context of the user and then measuring on the back end.

• 31:04Craig S. Smith?

  Is Ireland learning as as you operate within its its environment? Learning your your your workflows and and as he said, then suggesting new. Workflows or more efficient workflows?

• 31:20Speaker 01

  Yeah, listen, I if you think about the the AI universe in general. One of the core things, and you you can hear this from many of the founders of the various providers, many of the core exec teams. They talk about their building, you said one provider in particular, they were just right out and said it. They're learning the profile of the end user. They're learning their workflows of the end user, but they're not focused on what I just talked about. They're focused on targeted advertising. They they it was it was stated just directly. And that's why you saw, by the way, in the Super Bowl ads, I thought it was really funny, actually, when Anthropic had those cute Super Bowl ads that they were all talking about the advertising. It was really funny. But they're that world's all about that in the first place.

• 32:01Speaker 01

  We're just taking it using that type of effort for for an empowerment capability for the end user to be effective at their job. And at the end of the day, again, always within the guardrails, by the way, also always as well respecting the privacy of the end user, making sure that we don't run off skew of privacy concerns or issues there. So empowering that for the right audiences of users based on the right contextual situation. The user somewhere in a given geo around the world that that is not acceptable. This is everything we do in the world of Ireland from an audit logging and from an from an information understanding is is governed by context. It's not some fixed level of audit logging that's verbose all the time.

• 32:43Speaker 01

  You know, this audit law, this audit policy is contextually driven. So the US user, we do this for them. But someone else in a different part of the world, we may anonymize something if we need to, or a different, a different application engagement. Anonymize that one versus capture screenshots on this one. So it's very, it's delicate to meet the needs of the privacy folks that are, that have those concerns and those privacy mandates they have to live up to, too.

• 33:04Craig S. Smith?

  So with these new, new

• 33:09Craig S. Smith?

  products and new capabilities, what, what's, where do you see this going? Is, do you think that

• 33:22Craig S. Smith?

  that you're, you're helping enterprises move deeper into AI and, and AI agent workflows? Or is it

• 33:36Craig S. Smith?

  that, that it's a problem that exists today that will eventually be taken care of by AI itself or different agentic platforms? Will they build this in or will they adopt Ireland as an API layer? Yeah. So that it's, it's not being applied by the enterprise. Yeah, I'm just, how does, how do you see this developing?

• 34:02Speaker 01

  Well, it's important to point out that. Organizations are going to be multi provider organizations. It is, you know, and they're already that way. They don't, sometimes they don't even know it and they are. But the sanctioned universe of AI will come from a million different angles within the organization. You will have legal teams that want to use a given AI provider for that's specific to their universe. You will have medical practitioners. You know, you even saw recently, you know, what chat GPT health. That's a good example of that. Anthropic had some answers for that as well. But you see different, different needs for different providers and different models. They all have strengths. You know, you see a lot of development audiences naturally leaning into things like Claude, for example.

• 34:43Speaker 01

  Um, and so they'll be coming at the organization from 15 different angles, maybe even 20. They'll all have enterprise agreements across these things.

• 34:52Speaker 01

  Each one of them is going to have their own unique front door to leverage their, leverage their resources. What we want to do is we want to let the organization leverage the mechanics of the island policy. Build integration natively to those providers. And so when the user accesses any of those resources in their agentic workflows, in their generative work that they're doing and beyond, uh, we want to leverage in a single interface. Every single one of the providers simultaneously. That way you have one set of policies. You have one set of resources. You have one set of experiences for the end user and it's not the wild west anymore. You put some orchestration and a little bit of organization around it as you enable the end users.

• 35:34Speaker 01

  Uh, but I see that's where the, the future, the future holds because orgs are going to be signing all kinds of specialized resources. And they're going to have them at the end of the day, but you're not going to want, oh, the, my, my level of fidelity and policy is lacking over here. We all, that's, we're cool with that. Uh, and this one over here, we got really deep stuff. And because every one of the providers is all going to have differently. If you've gone into any of the AI providers as a consumer, just go to the settings. They're not all the same in the settings. Some of them have different settings and capabilities and things there. So, but you don't want non-uniformity at the end of the day, when you're an organization. Uh, you know, it becomes chaos from a manageability. It's just, it's, it's, it's unmanageable at the end of the day. You want manageability by having singular policies, singular sets of.

• 36:15Speaker 01

  Audits and other sets of stuff, but then, you know, harness the power of every one of those providers. And that's where I see the future is as orgs begin to adopt the next one, the next one, the next one. They just integrate it right into the mechanics of island. At the end of the day, now it's living just like the other providers are living on, and it's all the same confines and empowerment that the other providers are living in as well.

• 36:36Craig S. Smith?

  Yeah, can we pull back a little bit and give me a little bit of the history of of island and then the chronology of the different products. And and so we can see where you've moved to this most recent set of products. And then talk about where you see island going from here.

• 36:58Speaker 01

  Yeah, great question. So we think of everything at island not as a technological problem. We focused on the core use cases a customer has. And and then there's a set of mechanics that have to solve that use case. But philosophically speaking, we we use this phrase internally quite a bit. And, you know, hopefully it makes sense externally as well. We we talk about being tactically strategic. That sounds kind of weird to say, but let me explain a little bit.

• 37:24Speaker 01

  As an organization, you've got an initiative to solve a BYO problem. Well, someone says, all right, let's grab a team of people internally that are all smart and let's solve this BYO problem. And so you'll set up a stack of technologies, you'll set up processes around BYO. And you'll feel good about whatever that approach was, or maybe you won't feel so good. I don't know. It depends on what you selected, but you'll solve that. Someone else comes to the table and says, hey, we've got a merger and acquisition scenario. We we're, in fact, we're acquiring companies all the time. We've got deal rooms over here and you've got a, we've got, we've got these people we're acquiring. We've got to ship them devices and onboard them. Let's go figure out process for that and the stack of technologies for that. And that BYO initiative over here.

• 38:09Speaker 01

  In many cases, didn't do a whole lot to affect your, your outcome on this one. These were tactical, tactical. And then the next thing comes along, tactical again, next thing along. This is why you wind up with this bulk of stuff that are all disjointed because you solved the problem in a vacuum of anything that could be a strategic outcome. This is a very important thing in the universe of Ireland and the enterprise browser is, is we think of it being tactically strategic. The mechanics that we needed to solve contractor for data protection, for private access needs, for accessing internal apps built natively into the browser. They were the same mechanics that I use because the policy lives locally. If I've got now got to onboard somebody to a virtual deal room.

• 38:50Speaker 01

  It's an emergent acquisition, those same mechanics get applied there, but I might need these additional things. So we built a handful of additional capabilities and I'm not good through all the chronology of Ireland. But all it took was some incremental capabilities because you needed the data protection. You needed the contextual awareness, you needed the private access to the internal deal room resources. You just needed protected accounts, the ability to leverage injection of credentials for the user. And that was the next step for that use case. Oh, your next use case was let me eliminate my virtual desktop infrastructure. Start thinking about how I get rid of that kind of stuff because I don't love that experience. Well, think about contractors, you're onboarding the contractors today and you're often using all the things we did early. And your existing contractor world outside of Ireland, you're using VDI for that.

• 39:36Speaker 01

  So it took for us was incremental steps. And what I'm getting after is it lets the org tactically solve an initiative, but more with an eye toward a strategic outcome because the same exact architecture, the same exact mechanics lets you go tackle the next thing and then the next thing. And you just add an incremental thing. So back to your question you asked a moment ago. That leads us up to what happened yesterday with these announcements. These are for us incremental things that are really important, but they build upon the foundation of what we've already got and they fit naturally into that flow. So it's not some artificial bolt on that you see with, you know, an upstream cloud provider having to bolt on something to get actual visibility over prompts. Like it's not, you're looking at packets. You're not looking, you're not having, you don't have a presentation level point of view. So you're

• 40:20Speaker 01

  having to reassemble packets artificially for that kind of stuff. So for us, it's not, these aren't bolt ons, they're just natural next steps for that given use case that comes up. So the customer brings an X use case and we might find, you know, all these things were applicable. It requires two additional things and then we go construct that and now we have an additional use case, but it's problem centric. Again, with tactical design in mind, strategic outcome in mind, you know, after you solve the tactical issue. Does that make sense?

• 40:48Craig S. Smith?

  Yeah. Bradon, what would you think I should ask him here?

• 40:51Speaker 01

  Yeah. Well, I think certainly hitting on the status quo conversation, what are some of the status quo things people use today to try to assert their will in this AI universe?

• 41:05Speaker 01

  And I can go there and go hit a number of different areas that are just where the status quo fails. Like they're just not, they're just not assembled to live in the AI universe. And so, you know, the, the, the strategy from, from your, your status quo players is often, and I said a little bit about this earlier, but the status quo is it's a block page. I gotta block you from getting to personal stuff. That is not a strategy. That is not a strategy for enablement of the user. How do you enable the agentic workflow and make it a part of their fabric and to make recommendations of prompts and things like that? If you live in the network path, you don't have, you sit on the sideline watching this conversation.

• 41:44Craig S. Smith?

  I mean, do you want to talk more about that? Or that's why I was trying to get kind of a A map of of islands so that I I know where to focus.

• 41:56Speaker 01

  Well, here's why I was going there because that lets me lean into our future work on the network side of things. The Sassy Universe because because then I start re redefining how people leverage leverage network resources. And that's really important. The AI generation because orgs are going to have internal MCP resources. They're going to have their internal apps that are going to be a part of that fabric as well. And today, there's a big separation, a lot of separation between internal resources and the the universe of public AI resources, et cetera.

• 42:21Craig S. Smith?

  Yeah, so talk about that.

• 42:23Speaker 01

  So i- if you consider the the the status quo of how um you would you would layer your your existing technology stacks in an AI universe. Well, your your VDI infrastructure doesn't have a whole lot of play. Um, in fact, you know, there's some I'll almost maybe a little bit of irony or using AI in a VDI environment. You get the benefit of all the AI stuff and now living in this terrible experience. Um, your your upstream providers, they have a limited point of view on this because at the end of the day, it's not a packet conversation. Um, you know, you I I think of it this way, like, you know, you're not going to understand the disposition of a workflow by, you know, I think criminal investigations, like, you know, they come to a crime scene, they start picking up evidence

• 43:09Speaker 01

  and they try to forensically put the pieces of the puzzle together based on the evidence they see. That takes a little bit of time and it's effort and you can be wrong. You may not, you know, you might make mistakes. But wouldn't it be a lot better if you had a camera that just showed you everything? Um, but because you're sitting there in the network, you're reassembling stuff, try to reassemble the picture of what happened versus sitting literally at the screen of the user observing what happened. Um, this is why you don't see, you know, a lot of the networking technologies, these upstream will be abilities to search yourself. That's why they're sitting on the sideline in these conversations with their block pages. You got to block this and block that and block this. That's not AI strategy.

• 43:48Speaker 01

  Um, and then, you know, more more importantly as well is you got to think about how we modernize access to those resources where networking is important. Because you've got internal resources, orgs are gonna have their internal clusters. Many modern, many really advanced orgs have their own internal clusters. They can afford it, but they're gonna have apps and resources internally that that AI universe is gonna have to play with. Meaning it's gonna have to engage and let AI help you in that application engagement, whether it be for automation or whether it be for just summarizing the content of a series of application elements to give you, you know, glue these pictures together across these apps. But sometimes those things will be internal, so, and those things could be MCP resources as well. So rethinking how we instrument our path to access those apps, but still letting the presentation be

• 44:33Speaker 01

  the fundamental point of view where we, where we understand what the user is doing. Again, back to that camera analogy, you know, it's, it's not a, we're not trying to piece together the forensics here. I want to observe and how do you observe? You watch something and if I can watch what the user is doing, I can understand what workflows they should have. I could then understand they need to talk to these internal resources, et cetera. And then it's a very seamless experience again versus upstream trying to do some weird assertion of the unnatural in that case.

• 45:00Craig S. Smith?

  Yeah, you know, there's a lot of talk about with, with companies fielding products acting as customer zero or client zero. I presume you guys are an island user as well as the company providing it. Can you talk about, yeah.

• 45:20Speaker 01

  I do a demo for a customer of island. I show island using island. I, I show a lot of that. The, but I, but the good thing about what we're building and what we've, what we demonstrated yesterday, the actual technologies, we built the technology over the course of the past year. We've got customers already using it in production at scale. We just, we just, you know, essentially look at it this way. We kind of came out of stealth yesterday, but we already built the technology, you know, in this, in those areas. We've got customers using those technologies in GA environments and production large scale. So, you know, the, the cool thing is the things we announce, there's, there's a deliberate path in why we do it. But the, when we're announcing something, you're not usually customer zero at that point. Usually you're using something that's already been tried and true. We aim you at customers that are already getting value from those things.

• 46:05Speaker 01

  So all of that AI stuff yesterday is, uh, has got a path of customers using that stuff at scale and getting a lot of value out of it.

• 46:13Craig S. Smith?

  Yeah, and is there uh any uh difference in how a company of 200 employees uses this as opposed to a company of 2000 or more?

• 46:28Speaker 01

  I don't know if there's a difference in how they use it per se. I mean, certainly the use cases in a massive multinational org. They're going to be way more complex and way more broad and than a company of 200 users. So the problems they may have to solve uh could be different. Therefore, they may apply different mechanics, different policies of application. The actual way they use the technology doesn't differ a whole lot. But the uh the outcome i- is is different for each side of the house, meaning it's different, but it's actually not much different except for the scale. Um the multiple. So in massive global organization, you know, for the scale and size they are, you know, you imagine if you can reduce a

• 47:09Speaker 01

  a cluster of VDI infrastructure that you're spending 600 million dollars a year on maintaining hundreds of server racks and things like that. Uh you get a lot of value out of that. The economy, the economy of scale there. Um but a lot of small org probably doesn't have those racks and probably going to get that same value there, but they're going to get value in other areas. And by the way, they may get value and just reduce that one rack of of VDI for sure. So they've got and also the frustrating part of this, you know, because that's in those environments, you may have one IT person. And that IT person is now going to be the virtualization person. They've got to be the security person. They've got to be whatever. So uh that jack of all trades, we've got to make their job easier as well. So there's there's a lot of value in both sides of the fence. It's just uh but to your question, the the what happens in the browser, the mechanics, the tooling and resources.

• 47:53Speaker 01

  Uh you know, it's in the it's it's similar for the small org and the large org. At the end of the day, it's just the outcomes can be much grander for a larger.

• 48:01Craig S. Smith?

  Yeah, you were talking about uh how island uh proposes uh workflows and and other uh tools to the user uh after observing how the user is working. Uh does it do that for for policy as well? Does it review and enterprises? Policies and and then suggest, you know, say you you you don't have a policy that addresses this vulnerability or this risk. Does it do anything like that?

• 48:41Speaker 01

  Yeah, there's a there's actually some really cool capabilities. This is an opportunity for AI, by the way.

• 48:48Speaker 01

  You know, in the universe to start observing and understanding some of the something we we've been using for quite a while now is the ability to have what are called insights in the product where they're they're AI generated insights that look for the needle in the haystack. In some cases, they may look for configuration elements. In other cases, you know, it'd be great. You got a duplicate policy here. Why do you have a duplicate policy here? But they they essentially look for insights in a number of different areas. They look for insights in cybersecurity, in areas of digital experience, into user productivity, into data protection. So it gets in a lot of different areas. And what's happening is just bubbling up. You probably want to know about this right here. And you know, that that know about it could be, how do you go build a policy around the data protection because you didn't know this was happening?

• 49:33Speaker 01

  Or it could be just a you've got a misconfiguration over here on this on this part of your universe around MCP. You're throwing errors there and you didn't even know about it. So just heads up warning on that.

• 49:46Craig S. Smith?

  And who sees that? Because not every employee needs to see that.

• 49:51Speaker 01

  I could call out. So we communicate to everybody engaged in Ireland. Communication could be the end user for certain situations because that could be the state of the privacy that they're engaging because you're governing the presentation layer. So we can communicate to the end user all kinds of different things to make their job easier because sometimes they just don't know something. It would be great if a pop up came up on the screen and said, do this instead. So we communicate there. But back to your question is, you know, those insights are often communicated to the practitioners who are helping build policies around the given area. And good thing about it is you leverage our back at the end of the day. So different people have access to different perspectives based on the given needs. So your privacy audiences see the insights around privacy, the people that do data protection see the insights around that.

• 50:35Speaker 01

  Everybody gets their own part of the pie handled that that is their interest area and doesn't let them get overly assertive, but lets them again. There let their expertise be asserted in the place where it really truly belongs.

• 50:46Craig S. Smith?

  Does that mean that there needs to be kind of a manager watching what people are doing within Ireland? For example, of policy lapses to make sure that they're taken care of or or does it depend on the profile? Of the user does Ireland read that they they Ireland knows that? This is an IT manager who's in charge of policies, safety policies. Uh, so it'll surface those things to that manager or is there?

• 51:25Craig S. Smith?

  Uh, sort of an island super user that's watching everything and and, you know, directing. Uh, to different people in the organization.

• 51:36Speaker 01

  It it's a little bit of both. Um, you know, certainly there's somebody in any technology arena that you introduce into an organization. There's somebody that actually runs the stuff. Somebody that has to operate it. Um, and so certainly there's, you know, a lot of times there's the champions that saw the value in Ireland and they've got operational people that take it on. Um, and then, uh, there are, we we try to take the administrative burden out of a lot of this stuff. I mean, a big part of this is letting the humans. Work in the natural workflows of the organization. Uh, for example, I use a perfect one. Oftentimes in cybersecurity, we're lacking. You know, you've you've built a ticketing environment. You've got it. You you've invested hundreds. I don't know what you've invested. I don't even want to speak to the dollars, but you've got a ticketing environment like a service now or a Zendesk or whatever it is.

• 52:22Speaker 01

  You you've invested in all that. And then over here, you've got admins over here manually creating exceptions in cybersecurity. Somebody opens a ticket. And admin gets the notification, whoever the practitioner is for whatever dimension of the business they're in, and they got to go log into the system over here, manually create the exception. And then they go about their work. Two weeks later, five weeks later, six months later, and then they got six gazillion exceptions in place and they can't even remember why they're all there. This was the old, the old, um, the old, you know, you think about the old firewall rules. Well, I, I had orgs. I remember back in the day orgs telling me they had 10,000 plus firewall rules because they don't know what they're all doing. They don't know which ones, like if you kill them

• 53:03Speaker 01

  all, you may kill the business. So, um, but in the world of, of Ireland is, you know, we will also want to let the users engage the appropriate approval workflows. So tying naturally into service now, when a user bumps into a wall, you know, they bump into a page that you can't go here because policy dictates you can't go there, but you can, you can click, click a request button right there on the, right there in the browser, click a request button and it kicks off in a, you know, the ticketing screen right in the browser. You fill out the ticket and then all of a sudden it sends the message into the ticketing environment and it flows through the appropriate approvers, whatever stack that looks like. And when the final approval is done, it creates the exception naturally in the browser and the, in the policy, in the policy engine. And if you want to put time bound type of

• 53:45Speaker 01

  stuff like expired after two weeks, maybe I'm going to give a research team access to some unsanctioned part of the AI universe that we wouldn't normally give people to give them access for three weeks and automatically auto expire that warning the user, this is going to auto expire. You may need to recreate, you know, request an exception, but we try to do everything we can about your operational question. Do it in a way where the, the manual labor of the past is just not sitting there scaring everybody in the eyes of, of fundamentally having to operate this in the, in the way that you have in the past. One last little thing. We also think of, of constructs when you build policies, constructs that are very novel, that when you apply them, they have the greatest level of positive impact, but the lowest level of operational overhead. Think about DLP. DLP has been an operational

• 54:29Speaker 01

  washing machine for ages. People constantly tuning it all day long and tweaking it. Well, we just take a different approach to it. We, there's a starting point. We leverage an application boundary as a policy construct. The application boundary just says, here's a virtual perimeter where your corporate apps live. The browser lets user move data freely in that perimeter with the, within the appropriate policy, but the user can't take data and go over here to this tab that's personal Gmail and spill the data beyond the boundary into personal. It's a very novel construct, but we understand the disposition of the various things that are open in the browser and what the boundaries are. And as a result, I don't have to go through the minutia of constantly tweaking and tuning. We just try to think of back to your operational question. How do we operationalize this in a way where it's just lightweight on everybody that has

• 55:14Speaker 01

  to go through this and deal with it? And that sometimes comes in terms of the, you know, the people that run it every day, the people that sometimes have to get involved in the approval processes in the mix, leveraging AI at the right times in the process, and then just novel constructs as well.

• 55:28Craig S. Smith?

  Okay.