OTSP #9: The Uneven Introduction of AI to Employees

Companies are urging employees to adopt AI—but a software rollout, training session, or announcement doesn’t necessarily prepare people for changes to their work. When expectations keep shifting, and the destination remains uncertain, how can organizations build trust and help employees move forward? In episode 9 of On the Same Page, Shel and Mark explore AI adoption as a change management challenge and the role internal communicators should play. They discuss listening to employees, challenging leadership assumptions, and building a flexible approach to change—along with the gap between recognizing the need for change management skills and developing them. **Raw Transcript:** **Shel Holtz:** Hey, Mark, how’s your experience with AI these days? **Mark Dollins:** It’s never-ending. I keep getting both excited and frustrated with it, but it’s all about learning, right? **Shel Holtz:** It is. It’s about learning for a lot of employees as well. Many companies are still struggling to introduce AI and drive adoption in their organizations. I was reading about the Financial Times’ AI transformation program in 2024. They started by listening. They surveyed employees and received 725 responses. What they found was that the biggest obstacle was a lack of time, and that helped shape their approach. A lack of time isn’t resistance. When we’re asking people to adopt AI and integrate it into their workflows, we’re asking them to make a change. We tend to look for resistance, but not every challenge is resistance. How have you been thinking about change management and AI adoption, given your passion for change management and internal communications? **Mark Dollins:** Ironically, one thing communicators have increasingly been doing over the last several years is using AI to become more productive. It’s already saving us time. But I think professionals in this space increasingly understand that AI isn’t simply a technology rollout. It’s one of the biggest waves of change we need to step into, get in front of, and help manage. **Shel Holtz:** I guess the question starts with our role in managing this change. Are we taking the lead? Are we trying to get that coveted seat at the table so we can influence it? Are we waiting for the training and IT departments to say, “Here’s the program. Communicate this and drive the change”? What should we be doing? **Mark Dollins:** It’s a little of everything. In the research we’ve done—and we’ll talk more about this in another podcast—Jon Stemmle and I conduct an annual survey about communicators’ attitudes toward and uses of AI. We talk to a couple of hundred communicators around the world to understand what’s changing, what we’re working with, and what we’re struggling with. We’re finding that communicators are getting a bigger seat at the table. We’re having more influence in setting governance and policies, which is good. We’re using AI more, but we aren’t using it for more strategic purposes. The biggest thing we haven’t fully signed up for is treating AI as a change management initiative that we should help lead. We tend to view it as a technology rollout or a policy rollout connected to AI. But artificial intelligence is itself a massive change. We’re talking about seismic adjustments in how organizations work. That’s a space employee communicators should be stepping into, and some are beginning to do that. **Shel Holtz:** I absolutely agree. The question is whether we’re being asked to help design the change or simply sell decisions that have already been made. We should use the expertise we bring that others in the organization may not have to the same degree: identifying impacts on employees, testing leaders’ assumptions before a rollout, and listening. But responsibility requires authority. We can’t resolve workload, staffing, or incentive problems through messaging. What decisions must we be able to influence if we’re going to be held accountable, to any degree, for AI adoption? **Mark Dollins:** Even better, let’s not wait to be asked. Let’s step up and say, “We believe we can help this organization better manage this wave of change. Let’s figure out how.” Some numbers provide context for what I call moving upstream. In 2024, about 45% of communicators reported being involved in shaping AI usage in their organizations. That rose to 70% last year and 76% this year. That’s a good sign. We’re saying we’re shaping it, although I’m not exactly sure how that’s happening in every organization. Communicators aren’t just using new tools. We’re helping organizations explain, govern, and normalize new ways of working. That’s where change management comes in. We asked communicators whether they view change management as a competency they should have. That figure rose from 73% in 2024 to 81% last year and is at 79% this year. That tells us this isn’t being treated as a fad. It’s more of a structural expectation. We’re also seeing growing interest in training or certification to become more proficient at managing change through communications. Around 72–73% of this year’s respondents said AI requires significant change management communications. About 80% believe change management—which is different from change communications—is a competency they should acquire. And about 76% believe communicators should be involved in influencing AI use in their organizations. There’s clear evidence that, as a global community of employee communicators, we know what we should be doing. The real question is whether we’re doing what needs to be done. **Shel Holtz:** At some level, everything we communicate is change communication. We’re putting something out because we want something from people. We want them to do something new or different, stop doing something, do more of something, or think or believe something. All of that is change. But when it’s a software rollout, the scope may be more contained. We’re adopting Workday as our HR and payroll software, and everyone needs to start using it. I don’t know how much of that requires the full Prosci process or whatever change management protocols you follow. AI is different. What does the change look like at the end? One thing I learned when I began studying change communication—and I had a great mentor in Tudor Williams—was that you should be able to close your eyes and envision the outcome. I don’t think anybody can do that with AI yet. That presents an interesting challenge for a change program. **Mark Dollins:** AI can be a tool for some change management and communication tactics, but we still need to understand the disciplines themselves. The question we should ask is whether we really understand change management and how change communications works with it. Over the years, I’ve asked communicators that question, and I often hear, “Yes, I understand change communications. I put out organizational structure announcements, leadership changes, and executive updates.” Then I ask, “What else went with that?” Often, there’s silence. That doesn’t happen every time. Our community is getting better at this. But my experience is that many communicators don’t get into the specifics of the change management discipline. What do I mean by that? Understanding organizational readiness for change, for example. Is the organization ready to make this change? Who are the executive sponsors? Who are the champions who will help lead it? Do we understand the detractors and barriers? Do we have governance and processes for addressing barriers when the people involved can’t reach a resolution? Those are parts of change management that organizational communicators often aren’t taught when they learn how to communicate change. Whether the topic is digital transformation, cultural transformation, or AI adoption, it’s deeper than simply a technology change. These aren’t one-and-done activities. They require sustained, focused, strategic communications and an understanding of stakeholders’ individual needs. Not everybody experiences change in the same way, and that understanding should inform our communication. There’s far more to change management communications when it’s structured and aligned with the discipline of change management. Without that alignment, you can find yourself in a situation I’ve experienced: “Hey, Mark, we have a big change we want you to communicate.” I ask, “What’s the change management plan?” and get blank stares. Then someone says, “Why don’t you develop that?” That’s why, more than 10 or 15 years ago, I took the time to become certified as a change management professional. I chose Prosci, one of several options. It changed how I viewed the relationship between change communications and change management. The training took about four days and was tremendously helpful. I’m agnostic about which certification people choose. What matters is understanding the concepts we’ve discussed. That understanding helps us drive successful change through communications. How’s that for getting on my soapbox? **Shel Holtz:** That’s fine. I loved it. What we’ve seen in many organizations is almost the opposite of what you’re describing. Initially, companies asked communicators to tell employees to experiment with AI, without much clarity about the purpose. Then we told people to use AI a lot. We got into this whole “token maxing” thing, where people were using so much that they blew through a company’s annual budget in a month, and organizations had to dial it back. It’s hard to get people to buy into a change when you’re continually clawing back the messages you sent earlier. What are we asking employees to trust? What happens to the time they save? Will there be new expectations? Will roles change or disappear? Who gets development opportunities? We shouldn’t simply accept the message we’re asked to deliver. Part of change communication is pressing leaders to explain what’s been decided, what’s still uncertain, and how employees will be involved. Blanket assurances that AI won’t cost any jobs, or that it will completely transform the way we work, can make the job harder if we don’t have a plan, a vision, or an understanding of the outcomes. **Mark Dollins:** Completely agreed. There’s a difference between being a tactical communicator who says, “We’re rolling out AI. Use it in these three ways,” and asking, before the message goes out, “And then what? Where is this going? What’s the longer-term vision for AI in this organization?” I’ve had conversations with CEOs who have told me, “In less than a year, every role in our organization will have an AI component.” That’s fascinating, but how are you communicating it, and in what context? If every role has an AI component, what happens next? Will roles be redefined? Should employees expect that? A responsible communicator, working with the CEO, CHRO, and other leaders, should say, “We can’t communicate only one phase of this. We need to think more broadly.” It’s okay if some questions haven’t been answered. But let’s not promise that no jobs will ever go away because of AI. We can acknowledge that it’s possible while also explaining the opportunity for employees to learn technologies that expand their capabilities, wherever they are and wherever they go. Ideally, employees can help us redefine their roles within the organization. That doesn’t always happen, but it’s an outcome worth working toward. We need to be open about what we know and what we don’t: “We don’t know exactly where this technology will take us. Here’s where we are now, and here’s what we want you to explore so we can learn, improve, and help define the next chapter.” **Shel Holtz:** I read about a company that brought together 100 employees, I think in Nashville. They identified challenges they faced in their jobs, grouped people with similar challenges, and had them vibe-code AI solutions. Rather than look for existing tools to buy—or rent, as is often the case—they built solutions to their own problems. It was apparently successful. But I wondered whether it was a separate, one-off activity or part of a comprehensive strategy. Did employees know it was coming? Had they helped identify it as an approach they wanted to take? I think it’s great that the company did it. I don’t know whether it was part of a broader change strategy. But I worry about one-off activities that aren’t integrated into a process where employees can see progress toward an articulated goal. **Mark Dollins:** You’re describing exactly why change management creates value. It prompts those questions and discussions before communications go out. Ideally, in that situation, people would have discussed the direction and addressed the questions: “This is one step in a multistep process of defining our future through AI.” That’s where understanding change management comes into play. I’ll be candid: I don’t use every element of my training every time I help communicate a change. I put it in context. We may not need to go deeply into stakeholder assessment because much of that work has already been done. Or I may recognize that we don’t have executive sponsorship and that we need to address that weakness before moving forward. It’s about understanding all the components of successful change management. People sometimes hate hearing this, but communication should follow the change management plan. It shouldn’t have to define the whole plan. We should be able to say, “Here’s our plan for managing change. Now let’s determine how communications can best support it.” We can become better change management practitioners. With the right credentials and experience, we may even lead the change effort. But especially when we’re getting started, we’ll probably need a partner who knows more about the discipline and who sees us as a partner as well. Communication is the thread running across every element of change management. Without it, we’re done. **Shel Holtz:** Try making a change without communication and see how that goes. **Mark Dollins:** Without thoughtful communication. People communicate all the time, but without a strategic approach, their efforts can fail miserably. **Shel Holtz:** It’s important to get involved early enough. We’ve said that not every employee difficulty with AI is resistance, but there may well be resistance. You won’t discover what it is unless you ask. You can make assumptions, but I worry about relying on them. Our employee populations reflect the public to a large degree, and we know the range of public concerns about AI: the environmental impact of data centers, privacy and data security, even fears that it’s going to kill us all. You can’t address resistance without understanding it. If a sizable group of employees is likely to resist the way management plans to introduce AI, you need to be able to say, “We should rethink this approach.” That won’t happen if we’re brought in too late. **Mark Dollins:** Our survey also measured communicators’ skepticism about AI. Skepticism rose in 2026 after falling in 2025. Something happened between last year and this year that made us less trusting. About 45% of respondents said they remained worried about accuracy and privacy. If communicators have those concerns, we can be certain that many of our colleagues do as well. Resistance can have many sources, but as you said, you have to ask. **Shel Holtz:** You have to ask, and you may need to acknowledge that these issues exist across society, not just within our company. Explain how we’re addressing them, how we plan to use AI, and what we consider ethical adoption. I also think the plan needs to be a living, breathing thing. In many change programs, you know the outcome. With the Dow–DuPont merger, for example, you knew the endgame. We don’t know the endgame with AI. We don’t know exactly how workflows will change. Many companies haven’t even gotten that far. Many haven’t prepared their data so AI can retrieve information appropriately as part of a knowledge management system. This is a different animal. We need a sustainable, ongoing, flexible plan that can pivot as we get new information or frontier model companies release new capabilities. Many factors may require us to change direction. We need to do that strategically and consistently while maintaining employees’ trust that we know what we’re doing. **Mark Dollins:** Not to mention emerging regulation. The EU is ahead of the United States in this area. Accountability and regulatory frameworks will be another influence on how we use AI. When we talk about both struggling with AI and seeing its possibilities, we also have to recognize all the other influences taking shape. There’s one other research finding I wanted to mention. We looked at whether employee communicators—people with employee or internal communications in their titles or roles—were becoming certified in change management. We reviewed more than 500 public LinkedIn profiles and found a substantial disparity. Remember, 79% of our survey respondents said change management was a competency they needed. But in the LinkedIn profile sample, about 8% indicated that they were certified in change management. There was also a difference by seniority. We saw figures around 10–15% for VPs or heads of internal communications, compared with 2–4% for individual contributors with employee or internal communications in their titles. That suggests more senior communicators are recognizing the value of change management certification, while fewer people further down the talent pipeline are pursuing it. That concerns me. I’ll go back to your earlier question, Shel: What do we do that isn’t related to change? That’s particularly true of internal communications, but I’d make the argument for external communications too. A new product, service, or person involves change. Understanding what enables acceptance and the associated behavior feels like a natural part of our work. Why wouldn’t we want to understand that discipline if we’re going to be held accountable for communicating change and, in theory, the resulting behavior? It was an interesting finding from those LinkedIn profiles. **Shel Holtz:** It’s fascinating. You know I’m vice chair of the Global Communication Certification Council, and I’m trying to encourage communicators to become certified in communications. But, good heavens, you can have more than one set of letters after your name. They all provide value. I appreciate the time, Mark. **Mark Dollins:** Always fun, Shel. **Shel Holtz:** Listeners, we’d love to hear about your AI adoption, change communication, and change management journeys. Let us know. There were comments on our last episode, but I haven’t had a chance to pull them together. The next time we meet, we’ll discuss comments from that episode and this one before moving into our next topic. **Mark Dollins:** Great. We’d love that. **Shel Holtz:** See you then. **Mark Dollins:** All right. Take care. The post OTSP #9: The Uneven Introduction of AI to Employees appeared first on FIR Podcast Network.