Beyond AI Adoption: The Need for Structural Redesign

Jessica Ridella, IWF UK Member, IBM Global Technology Managing Director 

Over the past year, AI adoption has accelerated rapidly. Across industries, organisations are using it to support decisions, produce work, and move faster. But underneath all of that, the way work is structured has not changed at the same pace, and that is where the tension is visible. In my role working across large organisations, I see this play out frequently. 

There is real progress in what the technology can do, but day to day execution still depends on a small number of people to connect the dots, resolve ambiguity, and move things forward. We are scaling capability without redesigning how work happens.

Most organisations are still operating in an execution driven way, where work progresses because individuals’ step in, make decisions, and carry context across teams. It works, but it does not scale well. A design led system works differently. It builds clarity into the structure, so decisions do not need to keep escalating. The differences are clear- execution increases effort, while design improves judgment and reduces how often effort is needed.

Earlier in my career, I built my reputation on being reliable in complex situations, stepping in and solving problems. For a while, that worked. But over time, I noticed a pattern. The more capable you are, the more the system begins to depend on you. Decisions start to concentrate, and complexity gets absorbed rather than addressed.

Now, with AI, I am seeing a new version of the same dynamic. The people who used to unblock decisions are now being asked to review and validate AI generated outputs, just at a much higher volume. When the same person is needed multiple times to move something forward, the system is not actually scaling. In some cases, it is making it worse by formalising the bottleneck instead of removing it.

You can feel it when output increases, but decision making slows down. In many cases, AI is driving more production while also creating more checkpoints to manage it. It does not fix the underlying issue. Without changes to how work is structured, the same people simply end up being responsible for more.

The shift for me was moving from execution to design. Instead of asking what else I could take on, I started asking different questions. What decisions should not need to come to me? Where are we relying on individual effort instead of shared clarity? How do we make progress easier to repeat?

In practice, that means noticing where decisions consistently route back to the same people, making sure context travels with the work rather than sitting in someone’s head, and defining success not just by outcomes, but by how repeatable those outcomes are. If something only works because of one person, it is not a solution. It is a workaround.

This is especially relevant for women in leadership. Many of us have been rewarded for stepping in, fixing issues, and holding things together. Those are real strengths, but they often become the system’s safety net, relied on but rarely made visible.

In an AI enabled environment, the question is starting to shift. It is not how much we can carry; it is what we choose to redesign. Because if the system still depends on the same people to make it work, AI is not scaling the organization. It is just increasing the load.