Change management in the AI era: an interview with Tuomas Heroja

🇫🇮 This interview is available in Finnish (original); please find it using the language switcher on the top right corner.

How does AI transformation succeed and why does it often fail?

Recordly’s marketing lead Michaela Holmström sat down with data transformation coach Tuomas Heroja to discuss organizational change management in the context of the AI era. Their conversation explored why change triggers resistance, how organizations can evolve alongside AI, and what “human-centered AI transformation” really means.

Read Tuomas’s thoughts on AI, trust, and meaningful work below.

1. What does “change management” mean in the context of AI?

• - The basic idea of change management hasn’t changed, even though the technology around us has. Whether AI is involved or not, if we want something to change, it has to be led. Without leadership, results won’t materialize.

• The starting point for any work is always the same: we solve someone else’s problem. If we succeed, the customer pays us and determines whether the work was valuable. This doesn’t usually happen at a single point, it’s a chain. For example, if AI is introduced in customer service to help solve end-user problems, both the customer service agent’s behavior and the end-user’s experience change.

In that sense, change management isn’t different in the “AI era”. At its core, it’s always about people and changing behavior. But AI has introduced more fear into change. Public discussions about AI tend to be extreme, “everyone’s going to lose their jobs”, and that unnecessarily increases uncertainty. That’s why the quality of change management matters more than ever.

Often, organizations lack an understanding of what true change is based on: an individual’s own desire to change. Without a genuine connection to people’s experience and motivation, change won’t hold.

Motivation, however, requires involvement. It can’t be externally imposed; people need to have the opportunity to:

• understand what the change is about
• believe the change is the right thing to do
• believe it’s possible
• find their own role in making it happen (This idea comes from a former colleague of mine, Petri Aaltonen).

Too often, we see a big HR event where a leader explains the reasons for change from a stage, or a poster goes up saying “this is how we’re changing.” The assumption is that change happens on its own. Resistance in these cases is completely human and understandable, because people don’t feel heard.

In the end, the only way to solve this is through real dialogue. The goal is to build a productive business and that’s only possible through the people doing the actual work. The better we get at unlocking their potential and supporting intrinsic motivation, the better the decisions, the braver the experiments, and the more sustainable the productivity.

When AI becomes part of processes, it’s critical to start with a person’s core need: what’s the problem they’re facing, and what needs to change? Once that’s clear, we can step back and ask whether the change should be implemented using AI or something else.

It’s also important to distinguish two things:

• Change management using AI (the change itself can be anything)
• Integrating AI into existing processes

In conversation, we focus on the latter; what happens when AI is integrated into work. And in those situations, truly involving people is crucial. Change management is never just about processes. It’s about leading people, their energy, and their inspiration. And in the AI era, that’s more important than ever.

2. What’s the most common misconception companies have when implementing AI? Anything you’ve noticed from your client work?

Probably the most common misconception is thinking that AI will solve our problems easily without requiring us to change. That it’s a “plug and play” solution you can adopt and hope for the best. That you can outsource both problems and thinking to AI. In reality, that doesn’t work.

I’ve seen the same phenomenon before, for example during the rise of agile methods. There, too, people believed the process alone would solve problems; just have daily standups and Kanban boards, and you’re agile. But true agility, like successful AI adoption, is fundamentally about mindset and collaboration. It’s about the willingness to face uncertainty and explore what’s truly worth doing next - together.

AI doesn’t automatically invite that kind of conversation. In fact, it can steer us away from it if we’re not careful. Psychological safety, the ability to speak openly about uncertainty and ideas, is still the force that drives business results. AI doesn’t create that, but it needs it. Additionally, AI alone doesn’t create value; it requires an environment where it can work: quality data, shared understanding, and trust between people.

Another common issue is tech-centric thinking; overemphasizing a process that’s assumed to solve all problems. The focus is on the tool, not the customer. The question becomes “what can we do with this new tool?” instead of “what do our customers actually need, and how can we serve them better?” Not every problem requires AI, and some shouldn’t even involve it at all.

Often, I help clients return to a fundamental question: what is the problem we’re trying to solve? Only then can we assess if AI is the right approach. This usually happens through conversations in small groups, interviews, and building shared understanding of what’s important to us. Maybe this type of work isn’t that flashy, but it’s critical.

At its best, AI gives us new perspectives and helps clarify how things are connected. But it doesn’t replace human connection. I don’t believe AI could ever replace the moment when two people truly understand each other and are thereby able to create something unique together. 

AI is an excellent servant but a poor master. If we build our business solely around individual efficiency, we easily forget that real value only emerges when people’s ideas come together. That’s when something new can emerge; not just the same thing, done faster.

Think of a band and its audience. If there’s only an audience and no band - what happens? Or a band with no audience? We’re talking about an interaction that needs both parties. AI’s purpose is to connect people, not replace them. Replacement thinking is an extreme. And nothing valuable ever comes from extremes, except revolutions and trouble.

3. Could you share a situation where a team truly grew through an AI-related transformation? What made it successful?

I would describe one successful AI transformation through an analogy from a dog park. People come there with their dogs, but soon find themselves chatting with each other. The pets connect them, and that creates a foundation for new interaction. In this context, AI can be a bit like the dog; it brings people together to talk about things they might not otherwise dare or remember to discuss.

One key factor in success was that the team could experiment quickly without an exaggerated need for planning ahead.

AI brought courage to the table: they could try out ideas with a low threshold and learn fast what worked and what didn’t. We found a problem worth solving faster than before, but more importantly, trust within the team grew.

Similar to Peter M. Senge, I say: “: “Learning is a process of finding what is truly worth doing and how to get it done.”

In working life, there’s a certain amount of acting by default, almost as if we were actors on a stage. We say yes when we’d rather say no, driven by a need to please or a fear of upsetting someone. I’d like to reduce that by fostering genuine interaction between people, so that we can face even difficult issues in a positive and constructive way and find ways to move forward together.

We have a fun example of this in our own organization: our famous Thursday quizzes. They almost seemed to appear on their own. People joined in simply because they wanted to. That’s the most powerful kind of change - the kind that arises organically. It carries energy, curiosity, and a sense of shared purpose. Everyone gets to be fully themselves while also feeling part of something greater, their community.

4. What do you wish people experienced after working with AI transformation?

I wish people would go home at the end of the workday feeling energized. That they feel they did something meaningful; not just moved tickets in a system or optimized reports, but contributed to something of value, even at a global scale.

We have the freedom to think independently within the organization and we should use AI for something bigger than memes: for the things that actually matter. 

I also hope people come to understand that AI is just one technology among many. It wasn’t built to solve the deepest interpersonal issues, like trust. It doesn’t remove the need for interaction, nor should it. But it can help us learn faster what to focus on and what to let go of.

In the end, it’s always about solving peoples’ problems. That’s not fluff, it’s the essence of work.

5. What does a healthy, human-centered AI transformation look like from your perspective?

A healthy, human-centered AI transformation always begins with the person and with a real problem that’s truly worth solving, with or without AI. In my experience, organizations still have a lot to learn when it comes to identifying what really matters. This isn’t a one-time event but a continuous journey from the unknown to the known, and AI can help speed up that journey. That’s what healthy and sustainable change is all about.

The people who use AI must be involved throughout the entire process. It’s not about automating one step with AI unless it genuinely impacts the whole. Instead of sub-optimization, the focus should be on what brings value from end to end. Genuine collaboration between the people who complete the work is essential. AI can act as a conversation starter, much like how dogs connect people in parks.

It’s important that AI does not remain just a “nice to have.” It must be a meaningful part of the process that helps us understand what the most important work is right now. The lives of both companies and individuals are made up of decisions, some conscious, some not. AI can support us in making better decisions, but real learning only comes by hands-on doing.

A healthy AI transformation feels meaningful, and it affirms that the individual matters too.

6. If you could give one piece of advice to a leadership team about AI and change, what would it be?

Don’t assume that one single person knows everything. Involve others from the very beginning and focus on identifying the real, worthwhile problem to solve.

The leadership team doesn’t need to have all the answers. Instead, it should focus on building shared understanding. What is the concrete challenge that, if solved, would create the most value right now? You usually don’t get there with PowerPoint or slogans, but by talking to people whose work the change will impact.

Keep people genuinely at the center. Don’t start with technology, but with a problem that’s truly worth solving, whether AI is involved or not. Often the issue is not a lack of ideas, but knowing how to choose the most sensible one and move forward through experimentation. Many things are possible, but not everything is worth doing. Experimentation brings concreteness, and concreteness leads to understanding.

In the end, it’s about how we bring together different perspectives, experiences, and skills so we can make better decisions together. Leadership isn’t about trying to control everything yourself. It’s about supporting shared thinking to move forward.

7. Are traditional change management models still useful in the age of AI, or do we need something entirely new?

If by traditional change management we mean “management” in the sense of top-down control, measurement, and discipline, then that was never really what we needed and it doesn’t work any better in the AI era.

What we’ve always needed, and now especially, is leadership that helps people realize their potential. We need inspiration, energy, and direction. Not leadership that increases surveillance and fear, but one that creates meaning and strengthens trust between people.

AI doesn’t eliminate this need. On the contrary, it highlights it. When everything around us is changing quickly and new technologies are being introduced, we need leadership to keep us engaged. Not because it’s nice to have, but because that’s where business results are created.

So do we need something entirely new?

Maybe not completely new models, but we do need a completely new attitude toward how we view change. Not as a project, but as a continuous opportunity for growth, both individually and collectively.

8. What kind of mindset shifts do organizations need to adopt AI into their work?

The first and probably most important shift is moving away from traditional command-and-control leadership toward understanding people’s motivation and potential. Too often, organizations still operate through control, detailed upfront planning, and rigid accountability structures. But in AI-driven change, that’s not enough. The problems are complex, and our environment is in constant, rapid flux. What’s needed now is thinking that supports personal growth. 

AI isn’t just a technical add-on. At its best, it sharpens our focus and helps us fully grasp the present moment. It allows us to see more quickly whether the work we’re doing truly matters. Learning speeds up, because AI makes it easier to test and iterate ideas fast. And we should remember that we’re spending the best hours of our day on this work. At its best, AI helps make that time more meaningful, more impactful, and more productive.

Organizations need a mindset shift where work is no longer passed “over the fence” from one department to another. Instead, we need to work together to focus on the real problem and how to solve it. That requires ongoing dialogue, continuous learning, and sensitivity to where our energy can have the biggest impact right now.

Maybe organizations could start functioning more like symphony orchestras. Not everyone playing their own solo, but people tuning into both their individual role and the shared purpose of what they’re creating together in that exact moment. I believe that’s when something truly meaningful happens.

9. What kind of resistance do you encounter most often, and how do you help people deal with it?

The resistance I encounter most often isn’t really about AI itself. It usually comes from people not feeling genuinely heard. Whether the change involves AI or something else, resistance arises when people feel excluded. When someone else decides where things are headed and why, without their input.

This doesn’t mean that every opinion should automatically change the direction or the decisions. Involving people isn’t about trying to please everyone, it’s about genuinely wanting to understand different points of view. When people feel heard, we can move forward because they commit to the direction on their own terms.

I support people through change by creating space for open discussion, questions, and shared reflection. Often, it’s enough that someone listens with real intent and says out loud what many are thinking but don’t dare to say. That way, we stop resisting fear or uncertainty and start addressing them together. And that’s already a big step forward.

Summary: What did we learn about human-centered AI transformation?

• Change doesn’t happen without people. AI doesn’t remove the need for change management. In fact, it makes it even more important. People need to be genuinely involved and understand where things are headed and why
• AI isn’t a magic solution. It’s often assumed that AI can solve problems in a “plug and play” kind of way. In reality, we need shared thinking and experimentation to figure out what actually works
• Healthy change starts with a problem, not with technology. Companies should begin with real, meaningful problems worth solving and only then consider whether AI is the right tool
• Learning is speed up, but only through doing. AI can help us iterate solutions faster, shortening the journey from the unknown to the known. This requires the courage to experiment and discuss
• Humanity is a competitive advantage. Trust, psychological safety, and human interaction cannot be replaced by AI. When people collaborate, they uncover the most essential problems. And those become the foundation for sustainable AI transformation. That is, in practice, how productive business is created
• Leadership isn’t about knowing everything. It’s about direction and energy. Instead of control, we need inspiring, participatory leadership that helps people realize their potential
• The most common barrier is resistance, which arises when people don’t feel heard. Change is accepted when it feels meaningful and people have a chance to influence it themselves.

This isn’t a technical project. It’s a continuous journey toward what matters most to people and to the business right now.

Stay on top of the AI landscape 

If you want to be the first to know about new stories like this, subscribe to our newsletter.