You may still have the same title. You may still report to the same person. Your LinkedIn profile may say exactly what it said six months ago. The box around your name on the organization chart may not have moved. And yet the job may not be the same job anymore.
Maybe the team is smaller. Maybe you have more people reporting to you. Maybe responsibilities that used to belong somewhere else now belong to you. Maybe the budget changed. Maybe the organization reorganized around you. Maybe AI now handles part of the work that used to require hours of human effort. Maybe it does almost nothing useful in your work yet. Maybe expectations rose because someone believes it should.
Maybe your expertise is still valuable, but what people need from that expertise is changing. Maybe none of these changes arrived together. That may be the important part. We often talk about jobs as though they are relatively stable containers. There is a title, a description, a reporting relationship, a set of responsibilities, and an understood way of measuring success.
Then something changes. Usually we notice the obvious changes first. A promotion. A new boss. A restructuring. A layoff. A merger. A new technology. But some of the most consequential changes happen underneath those visible markers. The work changes. The expectations change.
The resources change. The boundaries change. The people change. The tools change. The decisions change. And sometimes the formal role catches up. Sometimes it does not. That raises the question at the heart of this perspective:
What is the job now?
The Field Read
There is no single story about work in 2026. That may be one of the most important things to understand. AI is spreading quickly, but not evenly. Organizations are restructuring, but not in the same ways. Some companies are reducing management layers. Others are asking managers to lead larger teams while continuing to carry individual work.
Some organizations are redesigning entire workflows around AI. Others are experimenting with a few tools. Many are somewhere in between. Workers doing similar jobs can now be having materially different experiences of the work itself. The evidence does not show one synchronized transformation. It shows different forces moving at different speeds.
That can make orientation difficult.
The manager's job is already moving
Gallup described organizations this year as "quietly rewriting the role of manager." Companies have been flattening structures and widening spans of control. Gallup's research found that 97% of managers have some individual-contributor responsibilities in addition to managing people, and managers spend a median 40% of their time on individual work.
As teams become larger, carrying both responsibilities becomes increasingly consequential. That is more than an organization-chart change. It changes the job. A manager responsible for five people while spending half the week doing individual work is doing one job. A manager responsible for fifteen people while carrying much of that same individual workload is doing something different, even if the title remains "manager." And this is occurring inside a broader environment of organizational pressure.
Gallup found that 63% of employees surveyed said workers had been asked to take on additional responsibilities, 58% reported reorganization or team restructuring, and 52% reported budget reductions. Among senior and executive leaders, 29% reported difficulty balancing people management with individual-contributor responsibilities, while 18% said the number of employees they managed had increased. Those numbers do not tell us what any particular leader should do. They tell us something more basic. The work inside a recognizable role can materially change.
AI is part of the story. It is not the whole story.
It would be easy to make this an AI issue. The evidence does not support doing that. AI adoption is real and growing. But depth of adoption varies enormously. A new NBER study using nationally representative U.S. Census Bureau data found that between November 2025 and January 2026, 18% of firms used AI in at least one business function.
Weighted by employment, the figure was 32%. Adoption was substantially higher among very large firms in information, professional services, and finance. But even among companies using AI, adoption was often narrow. Fifty-seven percent used AI in three or fewer business functions. Sixty-five percent limited worker-level AI use to three or fewer tasks.
Most firms reported using AI to augment work. Employment reductions associated directly with AI were rare. Another nationally representative NBER study published in August reached a similarly restrained conclusion. The researchers described current generative-AI use as widespread but shallow. AI is appearing across many occupations and tasks.
But within most individual tasks, fewer than half of workers use it. People doing very similar work can therefore be using AI very differently. That matters. One executive may work inside an organization where AI is already changing workflows, staffing assumptions, analysis, customer service, software development, and decision-making.
Another may work inside an organization where AI mostly means employees occasionally using a chatbot. Both can accurately say their company "uses AI." Those are not the same professional environment.
The skills inside jobs are moving too
PwC's 2026 Global AI Jobs Barometer analyzed more than one billion job advertisements across six continents. It found that skills requested in the most AI-exposed occupations are changing more than twice as quickly as skills in the least-exposed occupations. It also found that new tasks being added to AI-exposed roles were disproportionately likely to require capabilities such as judgment, empathy, creativity, and leadership.
That creates an important distinction. Technology can remove work from a job without removing the job. It can also add expectations. A finance professional may spend less time reconciling information and more time interpreting it.
A technical leader may spend less time personally producing something and more time evaluating what machines and people produce together. A manager may have more information available and still face harder judgment. A junior employee may gain access to capabilities previously available only through more experienced colleagues, while simultaneously being expected to exercise judgment earlier.
The technology does not simply subtract. It can rebundle the work. A 2026 NBER paper provides a useful way to think about this. An occupation's task list tells us what people do, but it does not explain why those tasks belong together as one job.
Technology can move the boundaries among tasks, rebundling jobs without necessarily eliminating them. That may be one of the quieter developments worth noticing. The job survives. The bundle changes.
Sometimes the organization changes after the work
Microsoft published an unusually candid account in September of what it had learned from hundreds of internal AI transformation efforts. One lesson was that inserting AI into an existing task often made the task faster without fundamentally changing the result. Larger gains appeared when teams reconsidered the workflow itself.
Microsoft's conclusion was not simply that organizations need better AI tools. It argued that people doing the work need a meaningful role in determining how the work changes because they understand where judgment matters, where processes fail, and where the actual work differs from its formal description. McKinsey's September research makes a similar distinction.
It argues that AI is beginning to affect how decisions are made, how work crosses organizational functions, which capabilities matter, and where value is created. Capturing that value therefore requires operating-model decisions, not simply technology deployment. A separate McKinsey analysis published days later described organizations as scaling AI agents faster than they are redesigning the work underneath them. Again, the important word may be underneath. The visible organization can remain recognizable while the work underneath it begins to move.
The new thing has arrived. The old thing has not disappeared.
This may be the pattern I find most useful. Work in 2026 does not look like a clean transition from one operating reality to another. The old thing often still works. The new thing is already here. Traditional organizational hierarchy still exists. Organizations are also flattening. Managers still manage people.
Many are also substantial individual contributors. Expertise still matters. AI can now perform pieces of work that previously required that expertise. Degrees still carry significant economic value. Skills and alternative pathways are becoming increasingly important. Human judgment remains central. Machines can increasingly produce analysis, code, documents, images, recommendations, and decisions that humans once produced themselves.
Organizations still write job descriptions. The actual work can move faster than those descriptions. This creates overlap. And overlap can be hard to interpret from inside the job. You may wonder whether you are falling behind when the expectation itself changed. You may think you are overloaded when the job has quietly absorbed work from two previous roles.
You may question your management ability when the span of control changed but the individual workload did not. You may assume AI should be transforming your work because everybody seems to be talking about it, while your organization has barely integrated it. Or you may find yourself working in a substantially AI-enabled environment while peers elsewhere in your profession are not.
None of those situations tells you what to do. They tell you that the first question may be more basic. What job am I actually doing now?
The Pattern
A pattern is beginning to earn its place here:
The work can change before the role does.
I would still hold that conclusion with some restraint. There is no evidence that every professional role is changing dramatically. There is no single force responsible for the changes that are occurring. And there is no common speed. That unevenness is part of the pattern. The evidence comes from several independent directions.
Organizations are restructuring. Managers are carrying different combinations of people leadership and individual work. AI adoption is spreading, but it remains substantially deeper in some firms, industries, and tasks than others. Skills inside AI-exposed occupations are changing unusually quickly. Organizations experimenting deeply with AI are discovering that technology adoption can require workflow, decision, capability, and operating-model redesign.
The ILO's review of empirical evidence reaches an appropriately restrained conclusion. Productivity gains from generative AI are real but uneven. Large-scale employment displacement remains limited. Some of the more important emerging consequences concern how work is organized, worker autonomy, coordination, and job quality.
Taken together, these signals do not establish that work has entered one universal transformation. They establish something more modest and perhaps more useful: The formal role may not be enough to tell you what the job has become.
Recognition From the Field
This condition may not feel dramatic. It may sound more like: "I'm still responsible for everything I was responsible for before. I just have fewer people." Or: "AI saves me time, but now more is expected because everyone knows it saves me time." Or: "I'm managing more people, but I haven't really stopped doing my old job."
Or: "The title is the same. I'm just not sure the work underneath it is." Or: "Everyone keeps talking about AI transformation. Honestly, we're barely using it." These are recognition statements, not research findings. But they help expose something the aggregate data can hide.
Two people with the same title can now be doing meaningfully different jobs. Two companies in the same industry can be operating at very different stages of technological change. Two managers with the same number of years of experience can face radically different expectations of what management now means. The title is useful information. It may no longer be sufficient information.
A Vessel in Changing Water
There is a reason navigation has remained such a useful metaphor for Executive Orientation. A vessel does not move in only one direction. Naval architecture describes six degrees of freedom: surge, sway, heave, roll, pitch, and yaw. A vessel can move forward while rolling. It can rise while turning.
It can be pushed sideways while maintaining forward motion. Several forces can act upon it simultaneously. Professional change can feel something like that. Not because scope equals pitch or technology equals yaw. There is no useful reason to manufacture that kind of one-to-one analogy. The simpler recognition is enough: Your job may not be changing in only one direction. Technology may be changing one part.
Staffing another. Expectations another. Authority another. Relationships another. Resources another. And those movements may not be synchronized. That is why simply asking, "Is my job changing?" may not be enough. A more useful question may be: What is changing inside the job, and what is still the same?
What Deserves Attention
These are not recommendations. They are places to look more closely.
- Notice the difference between the title and the work. What do you actually spend your time carrying now? What has been added, removed, automated, delegated, inherited, or quietly absorbed?
- Examine what changed around the job. Team size, budget, technology, leadership, customer expectations, organizational structure, and measures of success can alter the job without altering its title.
- Distinguish AI exposure from AI reality. What is actually changing in your work because of AI? What is merely being discussed, promised, piloted, or expected?
- Pay attention to the work that became more important when other work became easier. Automation does not necessarily remove judgment. Sometimes it increases the value of judgment.
- Ask whether the job and the system around it are changing at the same speed. A new workflow inside an old operating model can create a different condition from either the old or new system alone. The point is not to make every change consequential. The point is to see consequential change at the right size.
An Executive Orientation Lens
There is a temptation when a job begins feeling different to jump immediately to a career question. Should I stay? Should I leave? Do I need a different role? Am I falling behind? Should I learn AI? Should I become something else? Those questions may eventually matter. But Executive Orientation begins earlier.
The first task is recognition. If your job feels different, what changed? Was it you? Was it the work? Was it the organization? Was it the mandate? Was it the available support? Did responsibilities accumulate? Did authority change? Did technology remove part of the work and expose a different part? Did the job become broader?
Did the organization become leaner? Did expectations rise? Did something important remain exactly the same while everything around it moved? Several explanations may be true at once. That is why orientation precedes decision. A leader who concludes, "I can't do this anymore," may make one decision. A leader who concludes, "This organization is broken," may make another.
A leader who concludes, "AI is making my expertise obsolete," may make another. Any of those conclusions could eventually prove accurate. But each becomes dangerous when it arrives before the situation has been understood.
The earlier question is simpler:
What is the job now?
What We Would Not Conclude Yet
There are several conclusions I would deliberately resist. I would not conclude that AI is transforming everyone's job. Current adoption is too uneven for that. I would not conclude that AI is producing widespread job elimination. The empirical evidence currently shows much more augmentation and work reorganization than large-scale displacement.
I would not conclude that management is disappearing. Some organizations are flattening. That does not mean management itself is becoming unnecessary. Gallup's evidence instead suggests that the work of management may become more demanding when spans widen without corresponding changes in workload.
I would not conclude that every added responsibility represents organizational dysfunction. Roles evolve. Organizations adapt. Broader scope can be appropriate. I would not conclude that the job description is obsolete. It remains useful. It may simply be an incomplete description of the lived job. And I would not conclude that a materially changed job means someone should leave it. Recognition does not determine routing. It makes better routing possible.
Worth Your Time
- What Work Does Generative AI Do?
- The Microstructure of AI Diffusion
- 2026 Global AI Jobs Barometer
- Span of Control: What's the Optimal Team Size for Managers?
- The Key to AI Value Is Hiding in Plain Sight: Your Operating Model
- The Impact of GenAI on Jobs, Productivity and Work Organization
One Question to Carry Forward
If your title stayed the same but the work underneath it changed, what job are you actually doing now?
Quiet Continuation
The Executive Orientation Brief is a permanent collection of six research-backed perspectives for experienced leaders navigating changing roles, mandates, organizations, careers, authority, and professional conditions. Recognition comes before routing. The question is not yet what you should do about the job. First:
What is the job now?