AI Is Changing More Than Jobs. Are We Ready to Redesign Organizations?

Almost every conversation about the future of work eventually gets to AI.

What can we automate? Which tools should we adopt? Which activities will change? How much productivity can we gain?

These are reasonable questions. And clearly, important ones.

But more than 20 years of working in Human Capital and organizational transformation have taught me to look beyond the change itself and consider what happens to the organization around it.

I have worked through periods of growth, restructuring, organizational redesign and business transformation. One thing I have seen repeatedly is that we can change technology and processes much faster than we change the organization around them.

Roles remain. Management layers remain. Decision-making mechanisms remain. Sometimes, so do our expectations of managers.

With AI, I think this will become increasingly difficult to sustain.

Because if technology significantly changes how work gets done, we eventually have to ask:

How long can the organization in which that work takes place remain unchanged?

Perhaps we should stop starting with “How many people do we need?”

We define a role, establish its responsibilities and required capabilities, and then find the right people. It is a model we know well, and many of our workforce planning processes have been built around it.

AI is beginning to challenge that logic.

Look inside almost any role and you will find very different types of work. Some activities require experience, relationships and human judgment. In others, AI can work alongside people and significantly increase their capacity. And some activities may eventually be handled almost entirely by technology.

For me, this changes the workforce planning conversation.

The traditional question is:

“How many people do we need?”

Increasingly, I think we will need to start somewhere else:

“What are we trying to achieve, what capabilities do we need, and what is the right combination of people and technology to deliver it?”

That is more than a change in wording.

When we start with work and capabilities rather than the number of positions on an organization chart, the conversation changes. We begin to think differently about roles, skills, teams, productivity and structure.

I believe this will be one of the more significant organizational changes driven by AI.

And what, then, do we expect from managers?

This is a more difficult question.

In many organizations, a meaningful part of management work is still devoted to gathering information, consolidating reports, analysis, monitoring, coordination and moving information between different levels of the organization.

AI can already reduce a significant amount of this work.

I don't believe this makes managers less important.

But I do believe it forces us to become much clearer about the value we expect managers to create.

If information reaches the right people faster, if analysis can be completed in minutes, and if some coordination can be automated, managerial value shifts even further towards judgment and decision-making, prioritization, accountability, developing people and leading through situations where the data does not provide an obvious answer.

In other words, towards leadership in its real sense.

That is why I think we should also take an honest look at the management structures we have today.

How many management layers are genuinely necessary to lead the business, and how many exist largely to manage the flow of information?

I don't think the answer will be the same for every company.

Nor do I believe that AI automatically means fewer management layers.

But it gives us a very good reason to ask why each layer exists and what value it actually adds.

Sometimes we optimize before asking whether the process still makes sense

One thing I have encountered in organizational transformation is our natural tendency to start with what already exists.

It is understandable. Improving an existing process is easier than questioning whether that process should still exist in its current form.

AI makes this temptation even greater.

We can take an existing process, introduce AI at certain stages and quickly achieve a better result: less time, less manual work, perhaps lower cost.

Sometimes that is exactly the right thing to do.

But sometimes we risk becoming more efficient at something we would no longer design in the same way if we were starting from scratch today.

For me, this is where the distinction between optimization and transformation becomes important.

Before asking where AI can be introduced into a process, I think it is worth asking:

“If we were designing this process from scratch today, would we still build it this way?”

Perhaps we no longer need every approval.

Perhaps a decision can be made closer to where the problem actually occurs.

Perhaps information no longer needs to travel up and down several organizational layers.

Perhaps a role needs to be redesigned, rather than simply “supported” by AI.

And, just as importantly, we need to decide where we want human judgment and ultimate accountability to remain.

At that point, we are no longer talking only about implementing technology.

We are talking about how we want the organization to work.

I am not convinced that AI itself will be the competitive advantage

There is significant pressure to adopt AI today. Understandably so.

But I doubt that access to AI alone will be enough to differentiate a company over the long term.

The tools will spread. Costs will fall. Capabilities that seem extraordinary today will probably become commonplace.

If two companies have access to comparable technology, the difference will come from what they are able to do with it.

One company may use AI to reduce the time required for existing activities.

Another may use the same technology as the starting point for a much broader conversation: How should work be organized? Where should decisions be made? Which roles are still needed? What capabilities should we build? What should we expect from leaders?

Both companies will be able to say that they use AI.

But they may end up building very different organizations.

And I suspect that, over time, this is where the real difference will become visible.

If we were building the company today

After more than two decades of working with organizations through growth, restructuring and transformation, I have become cautious when I hear that a tool, a process or a new organization chart will “transform” an organization.

Transformation is rarely that simple.

More often, it requires us to reconsider things we have become accustomed to — things that may have worked very well for years: roles, structures, decision-making processes, capabilities and even what we expect leadership to mean.

AI does not make that part easier.

But it gives us a very good reason to have the conversation now.

So, in a discussion with a CEO or a board about AI, I would not stop at:

“Where can we implement AI?”

I would put another question on the table:

“If we were building the company today, from scratch, knowing what AI can do, would we still design the same organization?”

I don't think there is a universal answer.

But if the answer is no, then the AI conversation becomes much more interesting.

Because we are no longer talking only about technology or how many hours of work we can save.

We are talking about roles, structures, decisions, leadership and the capabilities we will need in the years ahead.

And ultimately, about what kind of organization we want to build around the new possibilities AI creates.