When AI becomes a colleague
How organizations can build agility by making change part of everyday work
Article | 2026-9-25
9 minute read
Executive Summary
For most of human history, humans have been the subject and technology the object. Agentic AI begins to overturn that structure.
As AI starts to perceive its environment, make decisions, and act with increasing autonomy, people themselves become part of what must be redesigned. Roles, skills, and ways of working are all subject to change. Organizations that see AI simply as an efficiency tool risk missing a larger opportunity for growth.
What organizations need now is “change agility,” the ability to make adaptation part of everyday operations through a continuous cycle of Sense, Decide, and Rebuild. Traditional change management, built around a predefined future state, is increasingly becoming a relic of a slower era.
The most important investment ahead is therefore not in AI technology alone. It is in the organization’s ability to continually reshape itself alongside its new AI colleagues. We call this capability enterprise crafting.
In the AI era, enduring competitive advantage will belong to people and organizations that can keep changing. So how can organizations build that agility before change outpaces them? This article explores what that takes in practice.
Why redesigning human roles has become a fundamentally different leadership agenda
Corporate transformation has always asked one question. How should we change? From IT modernization to digital transformation, one assumption has remained largely untouched. People are the actors, and technology is the object.
Agentic AI challenges that assumption. AI that can perceive, decide, and act autonomously is entering work traditionally dependent on human judgment. People may no longer occupy the unquestioned role of the sole “subject” of organizational transformation.
Borrowing from complexity theory, we are moving from a complicated world, where expertise can often lead to the right answer, to a complex one, where there may be no single correct answer. What works instead is trying something small, sensing what happens, and responding quickly.
So the conversation cannot stop at “How do we use AI?” Leaders also need to ask who makes decisions, where accountability sits, what people should own, and what they should stop doing.
AI is moving from tool to organizational participant
Until recently, AI was primarily a highly capable tool. Agentic AI is different.
By autonomously learning, making decisions, and taking action, it begins to function less like a tool and more like a colleague, contributing directly to how the organization gets work done.
This aligns with dynamic capabilities, the ability to sense change, seize opportunities, and reconfigure the organization. Agentic AI is beginning to share this role, contributing to the organization’s ability to adapt.
Within human teams, capabilities often multiply through interaction. People influence one another, complement one another, step aside when someone else is better positioned to act, and elevate each other’s performance. Yet when AI enters the picture, our thinking often becomes strangely mechanical.
Human capability + AI capability = additional value
That additive mindset is visible in the frequent discussion about moving people toward “higher-value work” through generative AI. AI can deliver significant efficiency gains, yet many leaders still struggle to define what higher-value work actually means within their organizations.
That should tell us something.
If AI is treated as an extra pair of hands, we may make the existing system faster without asking whether it should change. Once an AI colleague takes on part of a team’s capabilities, existing roles also need to be reconsidered. Adding AI without redefining responsibilities risks confusion rather than value.
The goal is for people and AI to contribute where they are strongest and strengthen the organization together (Figure 1).
When people themselves become part of the transformation
Historically, people have been the agents of transformation. With agentic AI, they remain agents of change, but also become part of what must be changed.
This is almost the reverse of traditional human-in-the-loop, where a person supervises an AI decision. With agentic AI, the person’s role and purpose must also evolve.
For people experiencing this change, the questions are not theoretical. If AI can do part of my work, where does my judgment matter most? What am I still responsible for? What should I stop doing? What new value am I expected to create? (Figure 2)
These changes affect careers and skills, but cannot be left to individuals alone.
Can conventional change management deal with this level of change? Our answer is no.
Why traditional change management is no longer enough
Traditional change management can be compared to an obstacle course. The organization starts at Point A, identifies a future state at Point B, and works through known barriers to arrive there. Frameworks such as Kotter’s eight-step model*¹ and the ADKAR model*² also assume a broadly linear view of change.
That assumption is difficult to sustain when the destination keeps moving. Technology organizations adapted earlier, moving from waterfall to agile and DevOps, with shorter cycles of experimentation and adjustment.
Organizational transformation has not evolved at the same pace. Many dominant change models retain essentially the same structure. That is why we describe traditional change management as a relic of an earlier era.
This does not mean change management has no value. It remains useful for distinct initiatives with a defined target state. But agentic AI is introducing continuous change. For people inside organizations, that can feel like one transformation starting before the previous one has ended. Increasingly, change itself is the normal state.
*¹ Kotter’s eight-step model: A framework developed by Harvard Business School professor John Kotter that sets out eight steps for leading successful organizational change.
*² ADKAR model: A goal-oriented change management model developed by Prosci, based on five elements required for individual change: Awareness, Desire, Knowledge, Ability, and Reinforcement.
Change agility as a core organizational capability
What organizations need instead is change agility: the underlying organizational fitness required to keep adapting even when the environment continues to change.
For the age of agentic AI, we define it through three processes (Figure 3):
1. Sense
Continuously detect signals of change from frontline operations, markets, and technology.
2. Decide
Make timely decisions using clear rules, decision rights, and delegated authority.
3. Rebuild
Test changes on a small scale and rapidly incorporate what is learned back into the organization.
Traditional change management resembles mountain climbing. Change is treated as a temporary period of stress, with everyone working toward the summit.
Change agility is closer to surfing. The waves do not stop. Organizations learn how to recognize them, respond to them, regain balance, and keep moving forward (Figure 4).
Organizations also need absorptive capacity, the ability to recognize new information, make sense of it, and apply it. Change agility has no final destination. That is why it must be continually exercised.
How to put change agility into practice
So how can organizations build change agility without creating yet another transformation program? The key is to make it part of everyday work.
Question 1: Where should we start?
As with ADKAR, the starting point is awareness, but it cannot remain a one-off event. Regular management meetings can include a standing question. What is changing around us? What are we noticing that we were not seeing a few months ago?
The aim is to make detection part of the normal operating rhythm, rather than waiting for the next transformation initiative.
Question 2: How should progress be measured?
Change agility is ongoing, so traditional before-and-after metrics are a poor fit. This can make return on investment difficult for leaders to see, while employees may feel that the effort never ends. The focus needs to shift from a static state to responsiveness.
Useful KPIs include the lead time from detecting a change signal to a management decision, and how often signals move from frontline teams to leadership. Put simply, how long does it take the organization to act on what the frontline notices?
Question 3: How should change agility be embedded into the organization?
Organizations should not create a new change-agility department or governance forum. That risks more bureaucracy, silos, and change fatigue. Instead, Sense, Decide, and Rebuild should be embedded into existing meetings, review cycles, and decision processes.
Consider one B2B equipment manufacturer we worked with. Signals from customer equipment and maintenance operations were not becoming organization-wide learning. Leadership initially asked for another transformation training program, but employees had already been through multiple initiatives. The mood was familiar. “Here we go again.”
The problem was not training. What was missing was a reliable way to turn what people noticed into decisions and action. We therefore declined to introduce another training program and redesigned the weekly review meeting instead.
A “change log” captured new signals, a decision-rights map clarified who could act, and priority issues were tested within two weeks. Decisions that had stalled under “take it away and review” became faster, while individual experiences became shared organizational learning.
The lesson is simple. Change agility is not about asking people to change more. It is about making it easier to notice change, decide, and act. That means redesigning the rhythm of work. In that sense, it is the agile transformation of organizational transformation itself.
Delegating decision rights can involve difficult trade-offs. Giving people authority means someone else has to let go of some control. An independent third party can help navigate those tensions and step away once the organization can sustain the rhythm.
Question 4: What future does change agility enable?
A future in which agentic AI plays a central role is not far away. Change agility is, in effect, the organization’s operating system for that future.
As AI takes on more responsibility, people will keep asking where their judgment matters most and what they should let go of. Human role redesign and change agility are two sides of the same coin. As roles change, the organization must change with them.
Uvance Wayfinders is therefore not focused simply on implementing AI technology. Our role is to help organizations build the fitness to keep changing alongside it.
Investment in AI capability matters. But competitive advantage may increasingly lie in people and organizations that continue to adapt, because that is what ultimately enables organizations to realize the full return on AI.
From transformation programs to enterprise crafting
Organizations are approaching a fundamentally different moment, one in which humans may no longer remain the unquestioned “subject” of transformation. Turning that change into opportunity will depend on redesigning human roles and building change agility.
As a consultant, I have seen many organizations remain “in transformation” for years. To me, that is not a sign of success.
In fact, I hope the term change agility itself will eventually disappear. Why? Because change should not remain a special project. It should become something organizations do almost unconsciously, as naturally as breathing. As long as we still need a special term for it, perhaps it has not yet become part of everyday organizational life.
Job crafting describes how individuals actively reshape the content and meaning of their work. What leaders now need may be the organizational equivalent: enterprise crafting (Figure 5).
In a constantly changing environment, organizations must continually reshape not only their roles and structures, but what they are here to do and why they matter.
Just as people craft their work, organizations need to learn to craft themselves. They, too, need to step away from being the sole “subject” of change and evolve alongside AI.
That, I believe, is the everyday future we should be creating with agentic AI as a new colleague.
Reimo Maruyama
Associate Partner, Lead of Strategy & Change Management, Uvance Wayfinders
After spending five and a half years in Research Sales within the Global Markets Division of an investment bank, she transitioned into the consulting industry. Following roles at both a strategy consulting firm and a Big Four consulting firm, she joined Fujitsu in September 2025.
Throughout her career, she has led numerous strategy development projects and served as the engagement lead on a wide range of initiatives. In addition to formulating strategy, she has focused on change management to ensure that strategic initiatives are translated into tangible business outcomes rather than remaining theoretical concepts. She currently leads Fujitsu's Strategy and Change Management team.
Uvance Wayfinders
Consulting by Fujitsu
Uvance Wayfinders combines AI-native consulting, deep industry expertise, and Fujitsu's technology capabilities to help organisations turn strategy into measurable business outcomes.
Related Information
Redefining leadership decisions in the AI age
Contact us