What Omnissa Learned From a 1000% Rise in Workplace AI Apps

[00:00:00] Your agents aren't producing accurate answers because they don't have a complete semantic understanding of your data. And Denodo is solving this and solving it through semantic consistency. Through semantic consistency, your agents can start making accurate predictions in real time. So see what else Denodo can do by visiting denodo.com to learn more. But now let me introduce you to today's guest.

[00:00:34] What happens when employees adopt AI faster than IT can govern it? Well, my guest today will argue that shadow AI isn't simply a security problem. It's actually a signal that people are finding new ways to work. And while many businesses are still figuring out exactly how to support them safely,

[00:00:55] my guest today will join me in a conversation that will cover a recently announced 1000% rise in AI assistant usage. And with that in mind, why blocking tools alone won't work. So no more saying, let's block AI, that will solve the problem. I think the toothpaste is officially out of the tube now.

[00:01:18] And also how IT can build autonomous workspaces that are self-configuring, self-healing and self-securing. So buckle up because it's time for me to beam your ears all the way to Mountain View where you can sit down with myself and today's guest as we talk about whether IT is helping employees run faster or simply tying their shoelaces together. But I've given you enough teasers and spoilers there.

[00:01:46] So it's time for me to officially introduce you to my guest right now. So thank you for joining me on the podcast today. Can you tell everyone listening a little about who you are and what you do? Of course. Thanks, Neil, for having me today. I'm Heyman Sahani. I'm Vice President of Product Management for Workspace ONE at Omnisa. I've been with Omnisa for over 10 years and I've been in the digital workspace industry for over 15.

[00:02:15] Well, thank you so much for taking the time to sit down with me today. There's a lot I want to talk about, especially I think because AI adoption inside organizations right now all across the world is accelerating at an incredible pace. But there's also that realization that it might not move this slow again. But your research suggests that many employees are choosing AI tools without IT approval, which is leading to the growing of shadow AI.

[00:02:43] And we've all heard that and or seen that in our news feeds recently. But what does it tell us about how people actually want to work? What would you take away when you see a stat like this? Yeah, Neil, this kind of reminds me of the iPhone moment, if you remember back in the day, right? The iPhones came out. Every large company I talked to, the enterprise IT team said, oh, these devices are never going to make it through the perimeter, right? And if you remember, the BlackBerry were the dominant devices.

[00:03:13] And I don't have to tell you where we are today. I feel like AI just had the smartphone moment, if you will, right? That same moment we experienced with the iPhone. Now, ChatGPT, when it came out, it showed everybody in the world how much more productive you can be in your personal lives by using Gen AI tools, right? And people took that and said, well, if I can be so much more productive in my personal life, why can't I use the same tools at work?

[00:03:39] And I think that's the reason for why we saw the 1,000% growth in Gen AI applications being used by people at work. Now, as you correctly mentioned, the State of the Digital Workspace Report did find something which I don't find it very surprising. But a lot of organizations today, they have sanctioned AI. Microsoft Copilot holds 90-plus percent of the sanctioned enterprise AI deployments, I think predominantly because of how it's bundled with M365.

[00:04:09] But if you look under the covers to actual usage on enterprise-managed iOS devices, 96% of AI usage is unsanctioned. And if you break into that, Copilot is only being used by 4%. ChatGPT on iOS is actually the highest usage, which sits at 9%, right? So, and Android tells you a completely different story.

[00:04:34] On Android, 99% is unsanctioned, where Gemini actually leads the way with 61% of the usage, of course, because Gemini is integrated into the Android operating system. So, what this sort of tells me is, you know, employees are not waiting for IT to define an AI strategy and waiting for them. It's kind of like, you know, when I got my iPhone in 2009, 2010, I brought it to work and started using it, whether my IT liked it or not. It's just what happened. And IT had to kind of figure it out.

[00:05:04] And I think we're in that same mode again, but with a trend that is so much bigger and moving so much faster. Yeah, I completely agree. And just going back to that 2007 moment when the iPhone came out, I still remember, I think it was Steve Ballmer from Microsoft, who was laughing at the phone saying, it doesn't even have a keyboard like a BlackBerry. Nobody's going to email on this thing. Do you remember that? Totally. Yeah. Yeah. And, you know, the keyboard was something, you know, I have to be honest, for a few weeks,

[00:05:32] I kind of missed the keyboard until you got used to the swipe typing and everything. But, yeah, it was great. Yeah. And, of course, fast forward to present day, many organizations instinctively respond by trying to block unauthorized AI applications, lock it down, shut it down. But, I mean, is this a realistic strategy? Or do IT leaders need a completely different mindset for supporting AI in the workplace? Because it's an old problem, really.

[00:06:02] You mentioned the digital disruption, the phone, the bringing your own device to work. And then there was also shadow IT, now shadow AI. It's the same problem repeating itself. But how should leaders approach this? Yeah. So, you know, we have the honor of working with a lot of regulated, highly regulated, brand-conscious, security-conscious enterprises. So, you know, I'll tell you it's not an easy problem. You know, shadow AI was not an easy problem. Shadow IT was not an easy problem.

[00:06:33] So, you know, everybody starts by blocking, as you said, right? And I see why if I'm a bank, if I'm a healthcare organization, it is my fiduciary duty to my shareholders to kind of do the right thing. If it's unsanctioned, I must block it. That's the policy. It's okay to do it, but you have to understand what happens if you do it and you kind of put a blind eye towards shadow AI, right? If you don't even know what's going on, what's going to happen is people will go around you.

[00:07:01] You won't even know what's happening. Innovation is going to slow down and your top talent is going to walk away because they want to be in companies that are using these types of tools. So, you know, the companies that we are working with, even if they have blocks in place, they're not turning a blind eye toward shadow AI. They still want to understand and get visibility into what is happening in the unsanctioned world.

[00:07:25] What are the types of users and personas and what are the types of use cases that they're using AI for? Then they drill in and see why is the sanctioned AI apps not useful enough for those particular personas and those particular use cases. They go have the conversation and figure out, hey, do we have to approve something different for these particular people for their use cases?

[00:07:49] Or do they just not understand how to use what we have provided and provide them better training, you know, which is more, again, persona specific and use case specific. So I would say, you know, less than a lot of people have learned from the smartphone moment and they're not quite making the same mistake of keep putting a blind eye towards, you know, shadow AI. So that's the good thing.

[00:08:13] And just to highlight the scale of what we're talking about here, one of the big stats in your report that was a thousand percent year over year increase in AI assistant usage. Now, beyond the obvious security concerns, what business opportunities could organizations be missing if they only view shadow AI as a compliance problem? Anything you see around this?

[00:08:39] Yeah, I mean, you know, one could argue that you're starting with smaller numbers for Gen AI applications a year ago. So a thousand percent is not that great. But we all know, you know, from our personal lives and everything we see that, you know, ChatGPT had a moment and this thousand percent year over year increase in AI assistant usage is pretty meaningful. But, you know, if a company takes a block only mindset and says, I'm just going to turn a blind eye towards shadow AI.

[00:09:06] All I provide is this one particular AI application, use it or lose it. I think they're going to miss out on a few different things. You know, one, they'll miss out on this opportunity to get workforce intelligence. You know, if you know, you know, certain persona of folks in your organization, let's call it your logistics workers or your retail workers that are on, you know, zebra devices doing a very specific use case.

[00:09:31] What they might need could be very different from what a creative designer sitting behind their desk might need from AI. So I think that workforce intelligence is really, really critical for people to understand so they can make these people, you know, 10x, 50x, 100x more productive with AI. The second thing I would say is competitive velocity. I believe with AI, everybody is going to deliver things better, cheaper, faster.

[00:09:58] And that velocity for companies that are saying, hey, we have figured out how to enable AI safely and let's go outpace our competitors. Those companies are going to take a big leap forward compared to ones that are sitting and having committee based decision makings on whether we should do this or not. Right. I kind of all this. Somebody mentioned this to me, you know, is IT in the game of helping people run fast or are they making them tie both their shoelaces for the two different shoes together?

[00:10:28] Right. Which which game is your IT organization in? And third, which I kind of mentioned before, is talent retention. You know, in the past, people used to just come and say, oh, OK, I'm starting. Which device are you going to give me? And, you know, here's your windows or here's whatever you want. Now it's like people ask at the time they're interviewing, OK, do I have device choice? I think the same is going to be true for AI. They're going to come and say, which AI tools do you use? How many tokens am I going to be allowed to use?

[00:10:56] What what happens if I exceed my token limits? What sort of token optimization tools do you guys have in place? I think all of those things are coming. And if people are still in the mode of, hey, should we use AI or not? Or this is the only AI we approve because that's what our finance team allowed us to do. They're going to find that, you know, business wise, they're going to lose the edge. And digital workspace is becoming more and more autonomous.

[00:11:23] And for anyone listening that may be struggling to visualize what that might look like in their world. What does an autonomous workspace look like in practice and how can possibly AI remove everyday friction without making employees feel like they're constantly being monitored? Because it feels like somewhat of a delicate balance sometimes. Yeah, it's a good question, right? So, I mean, you know, my Tesla drove me to work today. It's an autonomous vehicle.

[00:11:49] If it can self-drive on a road with so many different things, then, you know, a digital workspace is all digital. These are your laptops, your phones, your tablets. So, we kind of have this fundamental belief and mission that we should be able to take your digital workspace and make it self-configuring, self-healing, and self-securing. What does that mean? Let's start with self-configuring.

[00:12:12] Today, somebody in IT has to sit behind a console, turn a whole bunch of dials to make sure everything from your device lifecycle during onboarding to your business as usual to offboarding. You know, people have to keep tweaking those to make sure your devices are compliant, your devices are functional, you're having the best experience. Why not collect the right telemetry? Tell the system this is ideal state, and if anything changes, take me back to ideal state.

[00:12:38] So, a lot of things that, you know, we provide and the industry provides in general is allowing us to do that today from a self-configuring standpoint. Self-healing, you know, why should I ever have to go to an IT or any ticketing system and log a ticket? If this system detects there's a problem, automatically fix it. You know, prevent sparks from becoming wildfires. Like, I should never have to go open a ticket. I should not even have to think in terms of ticketing systems in the future, right?

[00:13:04] That's the concept of if you are successful with this mission of an autonomous workspace, that's what we mean by self-healing. And self-securing, you know, you hear so much about, you know, Mythos and Project Glasswing and what AI is going to do to vulnerabilities. And you look under the covers, most of these vulnerabilities are because these devices have poor hygiene, right? People have not been keeping them up to date. What if we can automate the whole idea of device hygiene through self-securing, right?

[00:13:34] That's the concept of what we mean by an autonomous workspace. And we believe, you know, to make it happen, you need this convergence of experience management and security. So you need a platform approach to do that well. And that's what, you know, we have been focused on. And that's why that's our core mission. An employee experience is something that often receives far less attention than security, especially right now.

[00:13:59] But your research highlighted how even small interruptions can have a long-lasting impact on our productivity. And if we go back before AI, I think there was a statistic with something that we're interrupted every 11 minutes upon average when inside an office. But how can organizations maybe use AI to create a smoother working day rather than simply introducing yet another application employees must have to learn, must have to get to grips with? Any way that you see improvement here?

[00:14:29] Yeah. I mean, you know, security budgets have been increasing, yet every company is facing, you know, bigger cybersecurity risks than before. Of course, nobody wants to go to the board and talk about a security incident. So, you know, the ROI and the justification is easy. But you go and ask a company, hey, how do you measure employee experience? And they tell you, oh, we ask a question in a survey. Okay, how many questions and how often? One question once a year. And the question is, how do you rate the IT experience?

[00:14:57] And that's how most organizations measure their employee experience, right? So, no, that's the wrong approach, right? You want to use the telemetry that you can collect from all these devices, which you already have. And again, this telemetry is not trying to say, hey, how often is this user taking breaks or anything like that? This is telemetry about the applications, about the operating system. How often are these apps operating systems crashing? You know, what are the things that are mismatching in terms of driver updates? How often is the app hanging and so forth, right?

[00:15:26] And what we found is, you know, Windows devices see three times poorer experience compared to Macs. And, you know, when we say three times, you know, what contributes to that? Seven and a half times more app hangs, three times more forced shutdowns. So, you know, let's take the example of a hang. If my computer is hung while I'm trying to perform an operation, I'm not calling IT and saying my computer was hung for whatever, 20, 30 seconds, right? I'm just silently absorbing that pain.

[00:15:53] And you multiply that across a 40,000 people company and that adds up quite a bit, right? And research shows it takes about 23 minutes for somebody to refocus on something after a disruption. So that's the hidden cost of, you know, not focusing on experience. And I think companies that have figured this out, they actually have departments for employee experience management. They're just not doing it with qualitative survey data, but with quantitative data using digital employee experience management tools.

[00:16:22] And AI is playing a huge part in this, right? Where you don't just collect the data, you use AI to figure out what are meaningful insights through anomaly detection for the ones that are meaningful. What's the root cause and how can you automate fixing it? So again, that's how a lot of that self-healing is playing into it. And I suspect that every single person listening is unified by having been stuck watching an app as it's hung for about 30, 40 seconds there. Seldom.

[00:16:50] Is there anything more frustrating than that? And it's usually right in the middle when you're trying to get something done as well. Or you're presenting something. You know, I call that the silent suffering. And outside of that, observability is thankfully becoming an increasingly important part of capability for IT teams. So on that side of things, how does better visibility help organizations understand how AI is being used?

[00:17:17] And also, how can those insights maybe better shape smarter decisions around everything from governance, training, and even technology investment? Yeah, I think that's a good question, right? If you're blind to what's going on, you're going to assume that whatever you have sanctioned is the right approach. And I think that's the mindset shift. If there's one thing I'd like your audience to take something from it is you can block things and you can provide sanctioned applications,

[00:17:44] but you still need visibility into what unsanctioned is being used because that visibility can give you tremendous insights about what people are craving for, where they're finding value from AI. And I'm not saying blindly go approve all those tools, right? Go do the right research. Make sure your legal team, your security team, your procurement teams, along with IT, are all in sync as you approve and sanction those resources. But it's not a one-size-fits-all.

[00:18:13] AI is not going to be a one-size-fits-all. Understand that and be open to what other things exist. On training, like I said, you might have to train based on personas. Your training is not going to be one-size-fits-all either. Your retail store associates may need different training than your logistics people in a warehouse,

[00:18:35] and that's different from people writing code that's a developer or people designing things in a graphical gaming studio, for example, right? So just understand that each of these needs, whether it's for what tools you sanction, what training you provide, is going to be custom. There's no one-size-fits-all. But most importantly, you know, also acknowledge that on technical investments, if you get those wrong,

[00:19:01] the price you're going to pay is going to be compounded right now because your competitors, if they get it right, are going to be moving at the speed of light. So the cost of moving slow is going to be a lot larger than it used to be in the past. And observability is not the end. It's a means to the end, but it kind of opens the door for you to see what is the art of the possible, what your users are craving for, and how you should make the right investments based on that.

[00:19:29] And at the very beginning of our conversation today, we're talking about that iPhone moment in 2007. And since then, the role of IT seems to have changed dramatically from, hey, here's a phone, a computer, a monitor, a keyboard, and mouse, and your login. We're the team that manages your devices and applications. Off you go, give us a call if you need anything. Now it's almost shaping the entire employee experience. So how do you see this role continuously evolving even more?

[00:19:58] And what new skills do you think IT leaders need as AI becomes part of everyday work and the businesses looking to them for help? Yeah. I mean, when I first started my career in IT at a financial institution, I remember they were fixed playbooks. And as long as I followed those playbooks and did my things with the change control board, everybody was happy. But the world is very different now, right? We're seeing this convergence of management, security, and experience.

[00:20:26] And that is putting IT sitting in the middle because they have to be the ones making all these changes. We see IT teams are shrinking or smaller. Their budgets are shrinking. But their role scope is increasing because they're kind of this cross-functional team that has needs coming from every part of the organization. To answer your question, I think there's a few different skills that I see the really good IT organizations pick up. You know, one, they're more data-driven IT organizations.

[00:20:55] They don't act based on, hey, let us, because we followed the process in the change control board, nobody asked the question of, did you collect the right data? It's just, did you follow the process? Those organizations are struggling. And the ones with the right data that are truly becoming data-driven IT organizations with people that not just look at the data, analyze the data, let the data kind of dictate the path they should go, they are really, really thriving right now. Second, you know, in the past,

[00:21:23] you could hire IT people that could sit behind a desk, do the work really well, and they would thrive. I don't think that's the persona of the IT people you could hire anymore, right? You need cross-functional fluency. These same people need to work with finance to justify ROI. They need to work with HR to justify productivity. They have to continue working with their IT counterparts, their security counterparts, and they are responsible for, ultimately, the employee experience because those are their customers, right?

[00:21:51] So this is not an inbound-only role anymore. This is a little bit more of an outbound role, and, you know, these people are bringing that fluency. And then, you know, from a pure digital workspace standpoint, in the past, IT was mostly in the business of just managing devices and device lifecycle management, managing identity of users. But the role is so much bigger now, right? You're balancing the needs of security, AI adoption, automation,

[00:22:18] all with smaller budgets and more role scope. So I don't think the definition is something you can write on a piece of paper and say, oh, you're not going to do anything outside of this anymore. So it's becoming a little bit like product management, if I may say that, right? You do what's needed to be done. You kind of have the buck stops here mentality, but you do it within the confines of the security guardrails, the compliance guardrails, and the industry that you're part of. Love that.

[00:22:48] And I always try and leave everybody listening with a few valuable takeaways. So if we have a listener or someone from an organization that is experiencing rapid AI adoption across their business and they want to avoid creating unnecessary risk or equally slowing down innovation, any practical steps that you would recommend that they take over the next, what, six to 12 months to maybe help build a better workplace

[00:23:13] where employees can use AI confidently, but also securely and productively? Any advice or tips there? Yeah. I mean, you know, first thing I would say is don't chase every AI tool, you know, focus on building the right foundation for your company, which involves picking the right set of tools, having the right set of data, understanding your personas, the use cases and having visibility into not only what you have sanctioned, but also the unsanctioned part.

[00:23:42] So you can learn as you go. And second, you know, create a cross-functional, you know, some organizations call it AI council, whatever you want to call the council, build teams that are looking at things from a line of business, IT, security, legal, procurement. All of these teams are coming together and making decisions and they're making decisions fast. You cannot operate slow in this market anymore.

[00:24:08] Sure. Be open to change, but at the same time, understand that, you know, if you send the wrong data to public LLMs, you have contractual obligations to your customers and, you know, you will get into legal trouble for it, right? So you cannot just say, yeah, I'm going to run as fast as possible chasing every AI tool because you have fiduciary duty towards your customers, your shareholders as well. So that would be my advice. And powerful advice too.

[00:24:37] And for anybody listening that is interested in talking a little more about how IT can move beyond simply just managing devices to designing an intelligent ecosystem that anticipates needs, remediates issues before they impact employees, or even have a look at some of the research that we referenced in our conversation today. Where would you like me to point everyone? I mean, they can all go to omnisa.com. You know, that's our public website.

[00:25:03] You can download the full digital workspace 2026 report. You can learn about our unified endpoint management products, our digital employee experience products, our security and compliance products. If they have, you know, further questions, I'm on X, I'm on LinkedIn. We have a pretty thriving community of, you know, folks, not just from Omnisa, but, you know, other customers that join that community. So a lot of times our customers like to learn from each other. You know, it doesn't strike the same way if I tell somebody to do something,

[00:25:33] then if they appear at another organization that has the exact same security needs, tell them this is what we're doing. So we kind of create those opportunities and I would encourage people to join the Omnisa community. Well, I love chatting with you today and I love the positivity when coming with working with AI, talking about enabling, not blocking. While the rise of shadow AI almost proves that the lockdown only approach is failing and not the way forward. But also on a positive side, the future of employee experience with AI.

[00:26:02] So I will include links to everything you mentioned there, including your LinkedIn, the research we've referenced and a few other things. So I encourage people to pop over to techtalksnetwork.com. They'll find a blog post associated with this episode with a useful link section. Check that out. Let's keep this conversation going. I'd love people listening to share their ideas, their experiences too. But more than anything, thank you for just starting this conversation today. Long may it continue. Thanks again. Thanks, Steve. Thanks for having me.

[00:26:36] I think having listened to my guest there, it's clear that the companies getting AI right aren't chasing every new tool or pretending shadow AI doesn't exist. They're the ones watching how their employees work, how they're learning and understanding where AI is creating value and then building the foundations to support it safely. And as IT takes greater responsibility for security, employee experience and AI adoption,

[00:27:04] it clearly shows just how much the IT department has evolved and how success will depend, will highly depend on visibility, faster decisions and ultimately understanding what people actually want and what people actually need to do their best work. So over to you. Is your IT team enabling employees to run faster or mainly tying their shoelaces together? I'd love to hear your thoughts on this.

[00:27:34] TechTalksNetwork.com. Remember, please connect with me on LinkedIn, just at Neil C. Hughes. You also get me on X and Instagram at Neil C. Hughes too. Many ways you can find me. But I'm afraid we're out of time today. I'll be back again tomorrow. Hopefully I can meet you in the same place and the same time. Speak with you tomorrow. Bye for now. Bye for now.