Why Enterprise AI Adoption Is Failing: AI Sprawl, Low Trust, and the ROI Problem

Speaker A: This episode is brought to you by Wacme, an SAP company. Wacmi's State of Digital Adoption 2026 surveyed 3,750 executives and employees at the same companies across 14 countries. 88% of leaders said their people have adequate tools. 21% of employees agreed same companies. Enterprise work stays fragmented across applications and, and that fragmentation compounds as AI advances. WalkMe guides people through change, brings AI into the flow of work, and shows executives how workflows actually perform. WalkMe turns AI potential into AI performance. The full study and WalkMe's new AI at Work Pulse Survey can both be found at walkme.com that's wme.com hello everyone.

Speaker B: Welcome to another episode of Future Ready Leadership. My guest today is Ophir Bloch. He's the SVP of corporate marketing over at WAK Me. Ophir, thanks for joining me.

Speaker C: Thanks for pronouncing my name properly, Jacob.

Speaker B: And yeah, I do my best. I do my best. Um, so we have a lot of different angles that we're going to talk about today as far as uh, you know, the state of adoption and with AI, what's happening, leadership, employees, all that sort of stuff. Um, why don't we start high level for people not familiar with the company. Maybe you can give a little context around what the company does and also the report so people can get a sense of who you surveyed, how many people, and then we can kind of jump into some of the stats. Sure.

Speaker C: So, uh, WACMI is, uh, an SAP company, has been for the past couple of years, uh, founded in 2011 and we basically pioneered uh, or invented a, uh, category, a technology category that we call digital adoption. And uh, we were the first digital adoption platform out there. And it really came from this idea that uh, instead of creating uh, tutorials and videos and tons of presentations and stuff like that, what if we could actually help people inside the applications themselves? Uh, so we provide a very robust platform using artificial intelligence on providing product analytics, uh, and then automation to guide people inside the flow of work. And then in the recent uh, years it's been kind of evolving into an AI adoption platform where we help organizations realize the full potential of their AI investments. Um, and to the second part of your question, this is a report that we've been running for the past five years. This is the fifth report that we've published, uh, the annual state of digital adoption, which also kind of transformed and shifted into more focus on AI obviously. And uh, what we try to do is uh, with these surveys, the state of digital adoption report and our AI Pulse surveys that we do, we try to understand the real kind of way that companies and mainly, uh, large organizations are deploying these tools, utilizing them, the difference, uh, in perception between management and how real work is actually executed. Um, so this, like I said, was our fifth report and we uncovered some really cool stuff that I assume we're going to be talking about today.

Speaker B: Yes, yes. M. High level though, what is your kind of general perception of what's going on in the AI world? And I know that's a very broad question and it can go in a lot of different angles, whether it's uh, kind of the backlash that we're seeing, whether it's the record investments, whether it's the ROI debate. But kind of if somebody were to ask you just what do you think of what's happening in the AI space? What are some of your kind of general impressions?

Speaker C: So it depends if you ask me or feel as an individual enthusiast, really, you know, I play around with AI all day, or if you ask, you know, what we're seeing with our customers and what actually Enterprise, um, A.I. looks like scale. And I'll try to combine, uh, both. Right. So as I said, I use AI daily and, and I follow the news and all the hype, et cetera. And I think that's what it is at this point. There is a lot of hype and expectations. Uh, but when it comes to actually how AI helps us execute work, um, you'll find that people are saying that they actually feel more frustrated or less secured or even maybe adding work on top of their work. Um, and m. I think the same things that we've seen with, uh, the different kind of waves of technology in the past, we're experiencing that again, there's different levels of knowledge in the workplace. There's different level of, you know, resistance to change, different generations. And with AI hitting every department and every application, we're also seeing kind of an increase in those kind of symptoms. Uh, Right. Where people, ah, are less familiar with AI or don't really want to, uh, embrace new ways of working. Um, so to sum it up, I think there's a lot of expectation and a lot of hype, but the reality is that we're. We're still not in a place where AI is really going to, you know, take our job. Jobs and do things for us. Um, I'd say that we're still in the kind of experimentation phase, um, slightly towards the starting to see value out of it, but still early days.

Speaker B: Okay. Yeah. And I would generally Agree with you as well. Um, there's certainly a lot of hype. I mean, the investments that organizations are making. I mean, you see AI in the headlines pretty much every single day, multiple times. Whether it's, uh, A.I. is going to be the end of the world, uh, whether people are debating. I was listening to one debate earlier today and I had to turn it off because I thought it was so silly. But somebody was saying, oh, by like 2040, you know, humans aren't even going to be around anymore. And I'm like, okay, I can't listen to this stuff. Uh, so you just see, like, some people believe it's going to be the end of the world, and other people are like, well, you know, we're still kind of early on. And other people. So the range of perceptions that people have, it's hard to figure out what, where is the reality, what is hype and what is fluff? And, uh, then, you know, you had people like Dario Amade, what was it a year ago, saying that by 2030, like 50% of white collar jobs are going to be automated. And that created massive panic. And now you see all these AI companies are walking it back and they're saying, no, no, no, no, actually it's going to be fine and we're going to have more work. So it's. Even the vendors themselves are not consistent with, with their own messaging.

Speaker C: Yeah. So look, um, I think both of us, we're not prophets, right? I don't think any of us can tell, uh, what the, especially not the workplace, right? What the workplace is going to look like in 10, 15 years from now. Um, but I can say this, and I like to quote Jeff, uh, Bezos on this, I actually saw he was interviewed. I don't remember the exact event, but he was interviewed and someone asked him, what is Amazon going to look like in 10 years from now? And he kind of got pissed. And he said, why does everyone want to know what's going to change with Amazon in the next 10 years? And no one's asking what's not going to change. And he said that that's what he built. That's the premise of how he built Amazon to be what it is, right? People are always gonna want good products, fast delivery, low cost. Right? And I think in our world, that constant is the human aspect. And look at what's happening in the workplace now, right? I think regardless of what change is brought on by technology, right? It was the metaverse a few years ago and was, uh, working from home, and then it was mobile uh, native first, et cetera. Every time there's this disruption that is broug on by technology, you still need people to be able to use it and you still need people to kind of change their behaviors and the ways of working. So I don't think that for the foreseeable future AI is going to replace us at work. Yes, there's going to be, uh, an evolution of how we work and certain aspects of work and it's fine, like we all need to upskill ourselves, but I don't think that we're going anywhere in the next, you know, let's be modest. At least five years from now.

Speaker B: Yeah, and maybe even longer. I mean, people forget that, um, you know, AI is not just a technology problem, but it's very much a human problem. Whether you look at governance, uh, you know, policies, restrictions, I mean even now when you go into some of these AI tools, they, they've been, I think stripped in some ways of their full capabilities for, for a number of reasons. Like if you go into cloud, the latest version, or even chat GPT and say, hey, find me a, you know, what's underrated stock that I should invest in right now. A lot of the times you'll get a response back from cloud that says, I'm sorry, I can't do this, I'm not a financial advisor. Right. So it, you already see that some of the capabilities are being, I don't know, safetied out of, of the platforms. So it's kind of like having a, a Ferrari but being told that you can't go over like 70 miles per hour of, you know, your speed limit is capped out. And so I think the more we kind of progress, the more we're going to see regulations and policies and guardrails. And so it seems to me that you kind of have a choice. You can either be fast or you can be safe. But being fast and safe does not seem like a realistic solution, uh, or approach for people. Right. You're going to, if you're going to be fast, you're going to break some things along the way. And so, um, this kind of ties into the report that you guys released as well, because so many people think, well, I'm just going to give everyone a subscription to Cloud or ChatGPT and then we're going to be done. But now you see so many of these human factors that are actually keeping, uh, the roi, the use cases, the value from being manifested. So maybe that's a good transition into, into the report and we can start off with just Maybe high level when you look at AI adoption across the board, uh, and maybe you can tell how many people did you guys survey? Any context around the report too. But when you look at AI adoption, what are you seeing right now? And maybe you can look at this in terms of the previous reports that you've done too, like what the trend line is.

Speaker C: Yeah. So, uh, like I said, this is an annual survey and uh, we survey over 3,000 individuals and we try to split it between the executives and then the actual kind of line of work. Um, and um, the trend has been that there's this kind of perception gap for the past three years, I believe, between how management perceive the changes and the availability of these tools, um, and the ability to actually see ROI out of them versus, you know, the folks that actually use these tools. Um, and I think this year that kind of heightened with AI, specifically where we saw that management was pretty confident with, yeah, we have the right tools in place and our people are, you know, know how to use them. And we're seeing great. Um, and there were also some data points about the number of tools that uh, you know, management thought that they had. And then when you look into the kind of actual work, uh, how these tools were being utilized, first you saw that there were way more tools than management thought that they had. Right. They just didn't have visibility into that. Um, and that's because it's just super easy to, you know, go and license these tools today. Right. You just pull out your card and license a $20 per month AI tool that does a very niche kind of, ah, job. Um, and because, you know, these tools were just, uh, purchased, implemented, nobody thought about adoption. Nobody thought about how do I actually drive a change in behavior and get people to actually use these. Um, how do I, you know, thinking about different cultures and different languages or different time zones, different skill sets, that was an afterthought. So what we learn is that deploying these tools on its own does not mean adoption. It doesn't equate to usage, and furthermore, it doesn't promise you, uh, utilization and actual ROI from these tools. That was very clear from the report.

Speaker B: Yeah, I mean the adoption one is interesting because we saw over the past few months, companies creating these leaderboards, the theme of token maxing became relevant and everybody was kind of tracking, well, you know, how many, uh, how many people are using AI on what day. And it's kind of a, I don't want to say completely useless metric, but it is not a very relevant metric to track any kind of value, uh, at all. And then you saw lots of organizations like Uber come out and say, well, we've. We've kind of blown through our entire AI and our token budget in a couple months for the whole year. And so there's just, I don't know, it feels like a lot of, um, chaos and uncertainty when it comes to the enterprise adoption. We actually saw this, I don't know if you remember way back in the day with these enterprise collaboration tools, uh, like, uh, Chatter and Yammer and Jive and Lithium technologies. It was the same thing back then too, where anybody, if they didn't like using like Chatter or Yammer, they could go find their own tool that they wanted, spend the, whatever, three to five dollars per user per month, and they deploy their own tools. Like, and you had so many different platforms inside the company, nothing was connected. You have like your official company tool and everyone's like, this is terrible. We're gonna go do our own thing. It kind of feels like that as well in, uh.

Speaker A: My point.

Speaker C: Yeah, that's exactly what I said. Right. It's, it's, we're experiencing the same wave of trans that we've been experiencing for the past 30 years. It's just a new technology and it became ubiquitous overnight. Right. So I think that is probably one of, uh, that's a bigger challenge to deal with. But you're right. I mean, the moment that organizations just assume that deploying a tool would change how people behave or change the outcome, that's the flawed approach. And the problem is that they're trying to solve this flawed approach with old methods, which is another thing that the report kind of surfaced. So, um, they're trying to deal with these new technologies with training and with creating videos and tutorials. But the pace of innovation and the rate that these tools are being introduced into the workplace are just nothing that, uh, uh, an average kind of employee can deal with. Right. And I think that kind of also amplifies the problem even more. You can't solve new problems with old tools.

Speaker B: Yeah, I mean, by the time you do a training video on, uh, you know, one, one version, uh, it takes weeks, probably longer, a month or two to create, uh, everything and roll it out. And at that point, especially if your company's using two, three, four different AI platforms, you already have, I don't know, two new models, new plugins, maybe a new company was introduced already. And it's kind of like you're always going to be behind. So it's just, it's just not possible to to catch up and keep up with that. So speaking of the adoption issue, um, the point that you made earlier is that simply tracking adoption is not, you know, it's kind of a vanity metric. Are there specific metrics that you think organizations should be tracking? Because this kind of ties into the ROI debate. And this is kind of the big, you know, the big topic of conversation in the enterprise right now is, where's the roi? Where's the value? Are, ah, we being more productive? Uh, so what are you seeing in that regard?

Speaker C: Yeah, I think it requires a mindset. It requires a change in the mindset, it requires a methodology, and sometimes, believe it or not, it requires a platform that can support you with that. So, um, I think first and foremost, organizations don't really have the visibility they need into their existing tools. And we spoke about it earlier. Every department and every individual can purchase their own tools today. Right. So I, you know, in marketing you can go and buy 50 tools and then finance has their tools and sales has their tools. Uh, the reality is that in a workplace, most of the workflows span multiple applications and cross multiple departments. So there's a lot of overlap and a lot of redundancy in the tools that we buy. Um, and then you have like, you know, the native AI tools and then you have the SaaS applications that kind of embedded AI on top of them. So it's, you know, we used to talk about SaaS sprawl. There's AI sprawl now. Right. So I think that the first thing that organizations need to do is to kind of tame the beast and kind of really get visibility into what they have. And there are tools for that different, uh, type of tool. We also have an offering around that. I'm not going to go into that, but I think that would be the first kind of thing, um, that I would do if I were, you know, running an IT organization. What is available, uh, within my tech stack? What are the capabilities that these tools have, who should have access to them? And start from there. Then I would go by defining business objectives, like what am I trying to do? And usually it typically comes down to I want to make money, I want to save money, or I want to mitigate risk, and what are the workflows attached to that and which applications support those desired workflows and outcomes. And from there I would track, uh, the utilization of these tools. So like we said, going in and out of a tool or utilize or using a license doesn't equate business outcome, doesn't equate roi. Um, you know, if I go in and out of a CRM all day, does that mean that I'm um, selling more or closing more business? Probably not. So what is the, what does good look like? What are the business objectives and goals? And then you need some sort of system to track that to see if you're actually able to move the needles. And what are the again, work backwards business objective workflows associated to that, applications that support that out of the existing portfolio and then decide if you need to supplement that with uh, other tools or platforms.

Speaker B: One of the things that I remember from the report which was pretty shocking was the spend the transformation budgets, I think it was 39 million, um, or thereabouts was kind of the average uh, digital transformation spend. It went up to 54.2 million. And the breakdown was that the majority of it, I think 59, almost 60% was going to AI related priorities, 35% to AI tools and 24% to governance, uh, and trust. And that's a very substantial line item budget increase for most organizations. We're not talking about like a couple thousand dollars here and there. I mean this is a huge, huge jump. And honestly I wouldn't be surprised if it kept increasing as more organizations realize that this is not just kind of a tech thing. Um, is that putting a lot of pressure are you finding on organizations? Because that's a pretty substantial jump in trying to get people to use these tools.

Speaker C: Yeah, look, if you track the reports, like I said, this is the fifth report. This number only increases uh, year on year. And um, recently with AI there's been a lot of pressure from investors and CEOs to become an AI niche native organization or AI first or transform into an AI kind of first organization. And that pressure and the competitive kind of climate, uh, that that's going on, uh, is forcing organizations to buy technology first and then think about, you know, how is this complementing my existing tech stack and what I need to do. So like I said, it creates a lot of false uh, um, expectations I guess. Um, and I think what we're going to see is to your point about token maxing, I think now we're going to see organization kind of taking a step back and saying, okay, I really need to cap this, I really need to have uh, some way to um, have visibility like I said earlier, uh, understand how is it being used, what am I trying to do with it. And then also I think the most important thing specifically in the context of AI is how do I share knowledge.

Speaker A: Right.

Speaker C: Because what's Happening now in the workplace is you have these pockets of learning, right? People are spending their weekends learning how to use these tools and bringing it back to work. But how do you share the knowledge? How do you make sure that it's accessible to everyone? Because that's the only true way that you can scale and see real, um, value from this. Because otherwise you're just going to have pockets of success. The ones that are forward looking, forward thinking and more advanced and other capabilities are just going to be better and then you're still going to have the laggard. So how do you make sure that everyone has access to the same tools, same knowledge and same capabilities?

Speaker B: Yeah, it's interesting because so I run a chro group, uh, called Future of Work Leaders. And we had a meeting, uh, in person a few months ago and one of the chros that was there was telling us the story. And she's from a massive organization which I won't name, but she was telling the story how they have some of their own internal AI tools and then one of the employees there, using agents or whatever, they built their own kind of like fun version of it. You know, they branded it with a, you know, a company name and just, just it's an AI tool but they made it fun. And over a period of several months they would notice that this thing would get massive adoption. And this is an organization, you know, hundreds of thousands of employees. And it got to a certain point where so many employees were using this and so many tokens were going to it that they actually had to throttle the AI usage on this particular tool and they had to not shut it down, but really limit to how much people are able to use it because it was, it ended up being a substantial, uh, line item cost. And more and more when I talk to organizations now, they say the same thing, that they're being very conscious. It almost seems like they're tracking spend more than the roi. It's like, you know, where's the cost going? If they will put that same level of diligence into tracking the value, you know, maybe you would be there. Um, but to your point, so many companies now are, you know, how much are we spending? What models are they using? Uh, because people are, you know, you're using probably the latest version of cloud to check the weather or to check a sports score, right? And it's, and sometimes I find myself doing that. Like if I have cloud open, I forget to change the model and I'm like, wait a minute, why am I using Cloud M5 to like ask a basic, uh, basic question.

Speaker C: Yeah.

Speaker B: And so you scale that across tens of thousands or hundreds of thousands of employees and you're like, wait a minute, what, where's the money going?

Speaker C: Yeah, look, that, that's going to have to change. Right? And, and I think that's something that actually you can, you can see that with Gen Z's, uh, more than, you know, other uh, generations in the workplace, where AI becomes more of an operating system than just an assistant, if that makes sense. It's, it's how they work. Um, and I think that, by the way, I'm the same, right? I'm in my cli, I am in the terminal all day and that's how I don't even open the ui. I do a lot of my work inside the terminal today. And I think we're going to increasingly see that type of behavior take place where the AI, uh, is beyond just, uh, being something that you consult with. It becomes a second brain, it becomes a colleague, it becomes your chief of staff. So you want to interact with it all day. Um, so that's where the more sophisticated ways of working, I think, needs to come into play. Like to your point, about changing models or thinking about the token or the tokens and the output, um, the question is if that is the expectations that we should have from our, uh, people in the workplace, or is that something that the company should enforce? There's probably somewhere in between. Right. Because, uh, I expect from my team to learn and how to be efficient in their work. But, but there's also expectation from the workplace on how to kind of enforce that and do more training on it. Yeah.

Speaker B: You also mentioned Gen Z, um, which has been quite interesting. And I think in your report you guys found that Gen Z is actually the Most comfortable with AI, what was it? 94.1%, but also the most likely to overstate their skills. 45% say they have pretended to be more skilled with AI than they are against 13% of baby boomers. And so with Gen Z, I think we're seeing something very, very interesting. There's so much of a, of a backlash. I've seen reports come out saying that Gen Z employees are actually sabotaging their AI use. Um, you see commencement speakers who are getting booed when they're speaking at colleges talking about data centers and AI. There's just a massive backlash, uh, specifically I think, coming from Gen Z. So I'm curious to get your thoughts on just kind of the Gen Z and AI usage, the backlash, and what your thoughts are on That

Speaker A: a quick word from our sponsor, WalkMe, an SAP company. WalkMe's new AI at Work Pulse survey found 90% of workers feel confident using AI, yet only a quarter say it works on the first try. Their flagship annual state of digital adoption, 2026 puts a cost on that. 51 working days a year lost to friction. Traditional training leaves organizations managing change one application at a time. WalkMe brings AI into the flow of work instead and it turns AI potential into AI performance. Learn more by going to walkme.com that's walkme.com yeah.

Speaker C: I wouldn't call it sabotage. You know, uh, Gen Z, they look like they're the most, uh, kind of ready, but I think it's because of, they have high confidence in themselves and the tools. It's native to them. Right. They kind of grew into it. Um, but I think they're also afraid to ask for help and it's kind of compensating for that, uh, in a way. So just this overconfident, uh, mood. Um, but look, I think what we're going to see in the workplace, like I said, we're going to see different generations and the Gen Zs are kind of growing natively into this environment and they see it differently um, than what maybe the older generation would see. And I'm trying to be politically correct here, um, but for them it's more than just a tool to get work done. It's literally an operating system. It's a life, ah, operating system sometimes. Right. Um, they see it as a psychologist, they see it as a friend, they consult on health issues, on everything. And I think it boosts their confidence because they say, yeah, I use it in my day to day, I know what this is. So of course I know how to use it. But in the workplace, in an enterprise environment, it's a completely different beast. Right. If you think about compliance and if you think about privacy and HIPAA and PII and stuff like that, um, and that's I think where the discrepancy is and I think that's where they're afraid to because if they raise their hand and say, you know what, actually I don't know how to integrate with our erp, that might look bad on them. So I think that overconfidence actually prevents them from asking for guidance and helps.

Speaker B: Yeah, there was um, I think this was on Yahoo Finance. I read this, this was earlier, this was a couple months ago. But the, the finding that I read it said something, um, 44 of Gen Z workers actively undermine their company's AI rollout. And, and again, this was one sample size. I don't think it was, um, you know, probably a couple thousand employees there as well. But I don't know, I found that very, very interesting that so many people, specifically Gen Z, are. You just don't. They view AI differently? And I'm, I'm assuming it's because when they were in college, we were all told that AI is bad. Right? Don't use AI for anything. And then all of a sudden you graduate from college, you get a job, and all of a sudden the organization says, we need you to use AI it's like, wait a minute, I was told for four years that AI is bad and it's evil and I can't go near it. And now you're telling me that I need to be AI fluent. It. So it's, you know, and then you, of course you see the doom and gloom scenario from people like Dario Amade, from Sam Altman. You know, obviously they're backtracking now, but when you hear all that, it's kind of like, well, no wonder we're seeing so much of a backlash. If you paint me this dystopian vision of a future, why would you assume that I want to live in that dystopian vision? Why don't you paint a positive, uh, vision, a positive outlook, uh, on how these things can, uh, unlock potential, give me more opportunity instead of just of the doom and gloom. So I think Gen Z is in, uh, an interesting spot.

Speaker C: Yeah. And, and I think it's on us as, as leaders to kind of understand, uh, these expectations and this reality, uh, and just, you know, be, be a better leader, uh, for our people on how to paint that, you know, better future. Uh, to your point, um, I don't think it'll be all doom and gloom. I think there's a, a bright future on that. Yeah. For this, uh, for this uh, generation. Um, but it's on us as leadership to see those opportunities and kind of guide the folks.

Speaker B: Where do you feel like trust is when it comes to AI? Do people trust the tools? There have been lots of situations where even fairly recently, I believe it was PwC, where they had some report that they were sending out and it was. They uh, used AI and it made up citations, it made up full reports.

Speaker C: Yeah.

Speaker B: You see court cases where, uh, big major law firms and criminal cases, they're submitting documents and briefs only to find that AI is just making up citations, making up court cases, making up stats that are not real. And so the trust Issue. It's kind of like, it works well a lot of the time, but it's that one or two times where it doesn't work, where, you know, it's just going to cause some, some massive damage or be embarrassing. Uh, you know, even with Claus, sometimes I've used it and I've noticed, um, occasionally it'll throw in like a Japanese character into a paragraph and I'm like, what's going on here? You know, I've seen it make up things. I've seen it occasionally create run on sentences. I've seen like, yeah, you, you want to trust it. And most of the time it does a good job. But every now and then you get something in there that you're like, wait a minute, uh, I can't use this.

Speaker C: So, you know, I think it's a combination of us learning to use these tools better. Uh, and part of it is also on the kind, uh, of Frontier Labs, uh, themselves, right? So if you take, for example, GPT 3.5 versus 5.6 SOL today, it sleeps and bounds better than what we had with GPT, uh, 3.5. It's more accurate, more reliable. Uh, you can actually instruct it not to add EM M dashes and not to, uh, hallucinate stuff. So we're already in the right direction. And we can only assume that GPT6 or Fable 7 or whatever the next Frontier model is, it'll be more accurate, more reliable, and, uh, provide better results. But on the other hand, I think it's on us as a society to say, this is an assistant. This is an augmentation of my capabilities. It's not here to replace my output or what I do. And I think it was Clay who recently created this, uh, manifesto of how to use, um, content that was generated with AI. You need to stand behind the content. You need to read everything. You need to be able to understand what the sources are of what you're writing and what you're putting in front of individuals. So it's on both of us. We, as, as individuals, need to better harness these tools, uh, and fact check and be accountable for the output. But on the other hand, I believe that we're going to see better outputs from the models themselves.

Speaker B: There's also an interesting gap between seniority levels, uh, and there have been a couple articles that are coming out. I know you have this in your report as well, where a lot of the people who are responsible for the budgets for, you know, getting the organization to use AI, those same people are not actually using the tools or not as heavily as they want everybody else to use it. And so you have situations where if you're a senior leader or a team leader, you tell everyone else, hey, use AI. We're an AI first organization. But then you yourself are not actively using it the way that you want everybody else to use it. And that to me is a very big challenge. Because it's one thing to tell people you want to be an AI first company, but employees emulate the behaviors of their leaders. So if you can't come forward and say, hey, here's how I as a leader have been using it. Here are some examples, here are some things that I've learned. Just telling everybody to use it is, I don't know, it seems kind of useless.

Speaker C: Yeah, absolutely.

Speaker B: Yeah.

Speaker C: And you know, it's, it's the same old, like, you know, lead by example. Um, same thing happened, by the way, during COVID when we came back to, you know, working from the office or hybrid work models. If a CEO found it important that their people work from the office, they also need to show up. If I want folks to come into the office two, three, four times a week, m. I as a leader need to be there as well. Um, and the same thing goes with this, uh, kind of evolution of technology. So I agree with you. And I think one of the things that we did see in the report was specifically around this kind of, you know, the, the, the trust or the confidence. Uh, we spoke about that gap between the executives and the employees. So 61% of executives in our report said that they trusted AI for work, but only 9% of our, of the employees said that they trusted AI.

Speaker B: Wow, 61 versus 9.

Speaker C: 61 versus 9. Yes. So, you know, the, um, you can say the majority of the executives trusted it for work versus nine. Because, um, you know, when you're. And I've seen examples, you know, I come from a marketing background and I've seen examples where people believe that AI can do a good, you know, marketing job. It can write content, it can write social media posts, it can produce images and videos and. Because maybe they experimented with it once and it produced a reasonable result. But when you really need to produce high quality work and you know, we spoke about accuracy, we spoke about the, you know, making sure that it's, uh, it's real data and not hallucinations. And you need to make my slight changes and stuff like that, that's where the, you know, these models fail. And the workers experienced that and felt that, um, and uh, and I think that was a Big gap that we saw. So, so yes, you'll see those differences between executives and employees, but it comes to, uh, how they use it or how often they use it, and for what capacity.

Speaker B: Hey, that's a, that's a massive, massive gap, um, in terms of confidence between executives and employees. Are you finding, like, why is that gap so low for employees? Do they don't trust it? Are they not getting reliable responses from AI? Do they think their company's monitoring them? Why are you finding that so few employees trust AI?

Speaker C: Wow. There are so many reasons. Um, um, the first one has more to do with psychology, right? And I think you, you've probably experienced this as well with your, uh, prob. With your smartphone, um, iPhone, whatever mobile, uh, device that you have. Uh, you download an application because you think it will do something that you want it to do, but then the reality is that it doesn't really do the thing that you wanted it to do. So what do you do with it? You either delete it or you leave it on your phone and never look back. And in many ways that's what happened with some of the early versions of AI in the workplace. Um, remember the early AI powered chatbots or stuff like that, they just, just didn't do the, they didn't deliver the expected outcome. Right? So we kind of went through this kind of hype cycle with them and then we either ditched them or never, you know, just left them there and never, uh, looked back. Uh, so that's part of the reason that's kind of our experience with these early versions of AI. Maybe you experimented with the early ChatGPT and you, you know, you said, it's crap, it can't write like I write. I'm never going to look back to it. Uh, the second reason is because of the, um, you know, applications that are being provisioned by the organ. Sometimes a, uh, company would make decisions based on commercial licenses or financial reasons or relationship or politics or whatever, um, and decide that they're going to opt into a certain tool because of a prior relationship that they have with a vendor, um, but it's not necessarily the right tool for the job that their employees need. So, um, you're either going to see this notion of shadow AI in the workplace where employees go and just buy their own tools because, because the tools that were provisioned don't deliver the job, or you're just gonna see resistance and underutilization of these tools because they don't fit the bill. Right. Um, and then last but not least, um, we touched that as well, maybe they're afraid, maybe they just don't want the tool to take their job. So I'm not gonna use this, I know how to write, I'm not gonna use this copilot to write my, uh, press releases or whatever. So it's a combination of all of those.

Speaker B: I guess I also wonder how much of it is trusting the tool versus trusting the company because, you know, I would imagine, and I don't know if there's any data on this, that some employees might be scared that the organization is either monitoring the prompts, looking at their usage. Uh, you know, I know, for example, at companies like Meta, and I think they've backtracked on this, but they were sending out memos saying that we're tracking your keystrokes, we're tracking, you know, we're taking screenshots of what's going on on your computer. And so in that case it's not so much I don't trust AI, it's like I don't trust you as a company. And so that to me is also, uh, probably, uh, a big issue. If you don't trust the company, it doesn't matter what tools they give you, you're not going to use them.

Speaker C: Yeah, I mean if, if your company gives you reasons not to trust it. Right, of course. Right. But that goes that, that goes to any aspect at work. It's not just about how to use AI. And I think, you know, companies that, that do a good job, uh, in this are companies that communicate. Right. And I think you can't over communicate something like this. You need to talk about the reasoning behind why you do what you do. Uh, here's what we're about to implement, here's why we're doing it, and then communicating it when you launch the tool. And at the end of the day there's a win. Win. Right. If an employee can do her job better and if it helps her deliver better outcome, shorter time, be more productive, then great, it's a win for her. And for the company, it's a win because, you know, now you, you're more productive as an organization, but you need to communicate that. And if a company has, you know, bad culture or, you know, it's just kind of a poisonous environment, I don't think it has to do with AI. It, you know, it applies to every aspect of work.

Speaker B: I agree. Um, another finding in there that I thought was pretty surprising is I think you found a third of enterprise, uh, employees or a third of the enterprise workforce has not engaged With AI tools and at all. Is that the correct stat? Cause I read that and I was like, oh, that seems massive that there's a third of organizations out there that are just kind of like. Claude.

Speaker C: What?

Speaker B: What? ChatGPT? Haven't heard of it, sorry.

Speaker C: But, well, believe it or not, it's, it's not just self reporting. We actually have the, the ability to, to monitor application, uh, usage. So, so it's, it's, it was cross kind of, uh, corroborated. And um. And yeah, I think, you know, again, it all goes back to the human factor. Uh, people resist change, people don't like when change happens to them. Uh, and if there's no clear what's in it for me, why bother? Why learn a new tool? Why add work on top of my work? Um, it may be misconceptions, but yes, uh, there are pockets in the organization where AI has still not been used. Um, I think we're probably going to see less of that in next year's report. Uh, and it's kind of diminishing, but still it's, it has to do with culture, with resistance to change, and with that fear of AI replacing me at work.

Speaker B: Yeah, I mean that to me is, uh, that's a massive, massive number

Speaker C: because,

Speaker B: um, we're, I mean that just shows how early we are in this entire, you know, AI enterprise adoption space if a third of companies are still not even using it.

Speaker C: Yeah, well, you have to remember also that not all of the, not, uh, all of the employees that were surveyed are necessarily knowledge workers. Right. So you can work on a factory floor operating machinery or computers. So, you know, you might be working for 150,000 employee organization, but you're just on the factory floor.

Speaker B: Okay, okay, got it. Yeah, that also makes sense there too. Um, what do you think makes successful. You mentioned this, I think earlier in the first, like two, three minutes when we were talking, that you can't just kind of take AI and bolt it onto things. So from the companies that are doing a good job, kind of extracting value, getting employees to use it, seeing success, legitimate use cases, what are they doing differently? And when you talk about kind of redesigning or thinking differently about work, what does that actually mean and what does that look like?

Speaker C: Yeah, I think there have been tools in the workplace that helped with automating workflows. We had technology like robotic process automation or BPMs and stuff like that. Um, I think the common problem was that we tried to take manual processes and workflows, sometimes even broken processes, and we tried to automate those. And I think that that's a flawed approach. Right? That's just like, that's a band Aid approach to how work gets. And I, and you know, I always say that, you know, imagine giving, um, the Internet to someone in the 1800s. What would they do with, uh, with the Internet? They would say, you know, I can send faster letters, I can, you know, maybe, uh, print stuff. But they would never think about Instagram. They would never think about, you know, Facebook, these things that we did with like web two.

Speaker B: Oh.

Speaker C: Or web 3.0. Right. So I think in many ways that's where we are with AI today. We're thinking about how to automate and digitize workflows and stuff that we do today. Um, and we're still automating broken processes and we're still kind of taking our manual ways of working and thinking about how AI can replace that, where in fact there's probably a better way to utilize it. Most organizations haven't uncovered that yet, but that to me is the right mindset that we should all embrace. Thinking about what are the net new things that AI enables us to do that we couldn't have done earlier. Right. That's the other side of the transformation coin. It's not just about, um, how can I optimize what I already have, but how can I actually transform and create net new opportunities for my organization.

Speaker B: What do you think of the organizations out there who are using AI as an excuse to let go of people? You know, you hear about this with Block. Uh, you know, so many organizations have been doing this. And then interestingly enough, you've also seen some companies bring employees back. You know, Ford, for example, when they let go of a lot of their master mechanics and engineers, they brought them back because they said, well, AI doesn't have the context that you have. Like, you've seen the evolution of engines and parts. And so it's not just about kind of looking at a product and trying to detect a defect, but kind of understanding the evolution, the process, you know, intuition, things like that. So to me, at least it seems very short sighted for organizations out there who are just cutting employees. I think Jensen Huang, he did an interview not that long ago from Nvidia, and he said that the companies who are doing this just lack imagination. Where instead you should be using AI to unlock new opportunities, go into new markets, create new products and services, deliver things faster. But you still have companies out there who are, uh, um, saying they're letting go of people and, and saying it's because of AI?

Speaker C: Yeah, yeah. And look, I think it's gonna, this is kind of the trend, um, that we're gonna be seeing in the next few years and it'll hit a convergence point. It'll hit a point where we find a balance between the human output and the benefits of AI. And that's where we're going to reach that convergence point. And um, you know, I think CEOs are under a lot of pressure. Uh, it's a very competitive landscape now. So uh, whether you're uh, privately held company or a public company, uh, there is a lot of pressure to show ROI from AI. And one of the easiest ways to show that is to say, yeah, I'm going to cut my workforce and then each employee is going to deliver 10x20x100x and that's how I benefit from uh, AI. But I think to your point, they're going to quickly realize that AI can't replace um, every knowledge worker and every uh, individual contributor and professionals in the workplace. And then we're going to try to upscale some, reskill others and uh, hire folks to do new jobs, new roles like AI architects or you know, uh, and I think some new jobs are going to emerge and it'll probably take a few years but we'll hit that convergence point and we'll find the balance.

Speaker B: Yeah. Ah, there's uh, the latest kind of trending. The forward deployed engineers that talent here had over years, uh, I think they had this job years ago. But now a lot of organizations are using this kind of same deployment where you basically take somebody experienced with AI, embed them on a team for several weeks or a month or two months and have them work with employees to actually understand the use cases, build things out and, and this, I mean I've seen some crazy salary ranges for these types of roles. Hundreds of thousands, four or five, $600,000 a year, uh, for these types of roles.

Speaker C: Yeah.

Speaker B: And then, and then, by the way, no one's talking about this.

Speaker C: Yeah, sorry to cut you off there. Um, I just wanted to add that on top of that you'll see that the more kind of offline, uh, or human to human and relationship kind of aspects are going to increasingly become a differentiator. So again, let's go back to the marketing kind of domain. Right. Uh, if everyone has the access to the same models and same output, same video models, images, et cetera, what really differentiates you is that human connection and your ability to stand out in a crowded room offline. Right. And that's a Different skill set. That's not something that AI can do. So you're, you're now seeing, or I think OpenAI recently hired a, ah, employee experience manager. Right. And, and other kind of roles that are more human relationship and offline functions.

Speaker B: Yeah, the, the value on kind of the human stuff is gonna, it's gonna be a huge, huge premium in that area.

Speaker C: Yeah, yeah.

Speaker A: Or so I believe.

Speaker B: And I think you believe the same thing as well. Um, because I think if everyone's using these AI tools, you know, collectively your organization might seem like it's getting smarter and more capable and the emails are better, the presentations are better, but you still need to be able to challenge people. And to me it seems like if you outsource a lot of your critical thinking, on paper you're smarter, but in person you're dumber. And, and that to me is a, you know, kind of an unintended potential consequence that organizations need to think about. Like if, if I know that you're submitting something to me and it's, I'm assuming you're going to use AI to get it done. I got to be able to pull things out. Right. You know, what did you think about this? What if you had to challenge AI, what would you challenge? What did you disagree? And being able to kind of extract that human discernment and decision making and creativity is I think, a massively underrated and probably one of the top, top, uh, leadership skills over the coming years.

Speaker C: Yeah, exactly. And, and like I said, this is going to be a transition phase. Um, but we're already getting better at it and I'm sure everyone can relate to this. Like, you know, back in the early GPT4O days, you just copied whatever the, the model kind of spat out and you pasted it in your emails. But now there are different tools and different ways to really think about it. We're more conscious about, you know, certain words, certain phrases, um, and embellishments.

Speaker B: Yeah, you can tell.

Speaker C: Yeah, you can tell. And I mean, so at least from what I'm observing at the moment, it is getting better. Um, and it's just a new capability, uh, that we're learning.

Speaker B: Yeah, and it's not only that, but now you're also seeing some platforms have AI detection tools. So for example, for me, one of the things that I do besides speaking and all that other stuff is I create a lot of content. Content. And so, you know, I have a substack, I put things up on LinkedIn and Substack recently introduced something where if somebody's reading your newsletter, there is an automatic kind of like, yeah, what you're reading was written by AI. Um, if you post on LinkedIn now, they're throttling AI generated content. So you can't just copy and paste from Claude and put it onto social media because no one's going to see it.

Speaker C: Yeah.

Speaker B: So I wonder if something like this is going to enter the enterprise at some point where when you're sending an email to somebody or presentation, there's going to be almost like a little, ah, a little icon that pops up that says, hey, just so you know, like 80% of this document that you're reading that Jacob put together on strategy is AI.

Speaker A: Right.

Speaker B: So. Well, I don't know if we're going to get to that, that angle or not.

Speaker C: No, I don't think that it matters because the other person is probably just giving your document to its agent and you know. Right. Uh, but I think, you know, the content aspect, writing, um, reading documents, the written word is just one very niche aspect of generative AI. And I think there's a world beyond that of workflow automation. And think about how pharma companies uh, are using AI to produce new drugs and come up with. So that has nothing to do with if the text was generated. Uh, it's more on the compute and the ability to kind of really extrapolate and think, uh, about solve really hard and difficult problems that we weren't able to solve with our Neanderthal minds. Um, and I think that's where the real benefit is. So not if someone wrote an email, uh, with the model or not.

Speaker B: Last question for you before we wrap up. Um, where's all this going? If we were having this conversation again in a year or in two years, do you think the world will be that different? Will the enterprise be that different? Or are we going to see kind of, you know, be roughly where we are now, but maybe just more people using it?

Speaker C: Yeah, look, I, I uh, really hope that the economic model, that economy will change. Um, I really hope that AI will solve real problems in the world. Um, you know, disease and uh, and different economics, tough economic situations. Um, you see all the guitars here behind me. I want a world where I have more time to play my guitars versus sitting in front of my computer or writing emails. And I think AI will allow us and kind of liberate us as a race, as a society to find more free time to spend it with our families, with our hobbies, with our passions, um, because we might not need five, uh, days a week to, to, to work for work. Um, and, and that's really the future that I'm looking forward to and I think we will reach that in the near, near term.

Speaker B: Yeah, that'll be, um, I think Elon Musk was one of the people that came out there and said, and you know, don't, don't have a savings account, don't save for retirement right now. You don't need it. You know, ah, we're going to have this massive world of abundance in the future. It's going to be this utopia. So we'll see. We'll see. Maybe that's the direction that we'll go. Yes, yes indeed. Time will tell. Um, Ofiya, where can people go to learn more about you? Wak me the report. Anything that you want to mention for people to check out?

Speaker C: Yeah, absolutely. So thanks for that. I'm uh, active on LinkedIn. We spoke about LinkedIn and I try not to produce AI slope. So, uh, feel free to follow me on LinkedIn. And uh, of course the company profile there, that's where we post most of our research and findings. Uh, you can subscribe to our newsletters where we always share, uh, first with our audience and customers. All of the great, uh, reports and uh, thought leadership that we put out there. Um, if you don't know where to start, just go to wachme.com and everything will be in front of you.

Speaker B: Very cool. Well, thank you so much for taking time out of your day walking, uh, us through this report, sharing some of the stats and your insights. Really appreciate it.

Speaker C: Absolutely. Thanks for having me.

Speaker A: Thanks to wacme, an SAP company, for supporting this episode. Walkme pioneered the world's leading digital adoption platform and today it is the platform that drives AI adoption across the enterprise. Their annual flagship State of Digital Adoption Report 2026 and the new AI at Work Pulse survey are both linked in the show notes. You can learn more by going to walkme.com that's walkme.com.