You can be an excellent driver and still be stuck in traffic. The skill is real and the car is fine, but neither one decides how long the journey takes. The road does.

That is a fair description of AI at work right now.

The people are ready. The workflows around them are not.

The Readiness Gap is Not a Training Problem

McKinsey surveyed 750 employees between February and April 2026. 70% said they felt personally prepared to adopt and use AI. Only 27% of leaders believed their organizations were ready for the shifts an agentic future requires.

The gap sharpens when you look at what actually drives results. Organizational readiness accounted for 48% of the difference between leaders who reported capturing value from AI and those who did not. Personal readiness accounted for 25%.

For roughly three years, most organizations have been fixing the person through prompt training, tool rollouts, and champion programs. That work succeeded, and people really are ready. The problem is that they are carrying all of that readiness back into a workflow nobody changed.

Same Tool, Two Very Different Experiences

Take the AI notetaker as an example, because almost everyone has access to one now.

In a typical weekly status meeting, the AI notetaker joins, transcribes, and produces a clean summary with action items. Genuinely useful.

Now… name one meeting that got shorter.

The notetaker made the meeting easier to remember. It did not make the meeting worth having.

Now redesign the workflow around the same tool.

1. Notes from the last session become a pre-meeting digest, circulated the day before.
2. The agenda covers decisions and disagreements only.
3. Attendance narrows to the people with decision rights.
4. Decisions are logged live instead of written up afterwards.
5. The recurring status slot disappears, because status is now read rather than performed.

Same tool, very different experiences. In the first version, one person got faster. In the second, the organization did.

McKinsey puts a number on that difference. Leaders were 5.3 times more likely to report enterprise value capture when workflows were redesigned than when they were left unchanged, 32% against 6%.

Where the Saved Time Actually Goes

The uncomfortable finding in the research is what happens to freed capacity. Employees gain personal efficiency, but that time does not automatically flow toward enterprise priorities. It often goes to work people personally find interesting, which is not always what the business needs most.

This is why most teams cannot answer three fairly basic questions:

• How many hours did AI give us back this quarter?
• Where did those hours actually go?
• What would we do with them if we chose on purpose?

Most people never got to the third question because they are stuck at the second.

Three Horizons, and Why Most Organizations Stall in the First

McKinsey groups AI maturity into three horizons.

1. Enablement, where employees use AI tools inside their existing jobs.
2. Automation, where end-to-end workflows across functions are improved at scale.
3. Reinvention, where roles, workflows, and operating models are redesigned with AI at the core.

46% of organizations sit in Enablement, 43% in Automation, and 11% in Reinvention. The value follows the maturity. 13% of leaders in Enablement report meaningful enterprise value, rising to 24% in Automation and 48% in Reinvention.

The report reaches for an old comparison to explain the stall. When factories first electrified, they swapped steam engines for electric motors and kept the same layouts, the same workflows, the same management systems. While electricity was obviously the better technology, productivity barely moved. Why? The gains arrived only once the factory itself was redesigned around what electricity made possible.

Even among organizations that reached Reinvention, 44% of leaders said they were still not ready for the people and culture shifts required. Maturity does not remove the human work. It exposes it.

What Welch Cut

In 1981, shortly after Jack Welch took over General Electric, Peter Drucker asked him two questions.

1. If you were not already in this business, would you enter it today?
2. And if the answer is no, what are you going to do about it?

Welch turned that into a rule. Every business would be first or second in its market, or it would be fixed, sold, or closed. GE grew from around $25 billion in revenue to roughly $130 billion over his twenty years.

The detail worth noticing is that plenty of what Welch cut was profitable. It simply would not have been chosen again. Drucker called the practice “Organized Abandonment”, and he had written the sharpest version of it years earlier.

There is surely nothing quite so useless as doing with great efficiency what should not be done at all. – Peter Drucker

Most of what clogs an organization is not failing. It was inherited, and nobody has been asked to defend it since. AI will now run every one of those workflows faster than before. That is the danger, not the benefit.

Ending Thoughts:

The readiness gap is not a story about employees falling behind. They are ahead. The organization is the part that has not moved, and no amount of additional tool training closes a gap that sits in the process rather than the person.

The good news is that the highest-leverage change in the research is also the most available one. You do not need to rebuild the operating model this quarter. You need one workflow, chosen because it genuinely matters, rebuilt around what AI now makes possible.

If your people are trained and the results still are not showing up, the workflow is usually where the answer is hiding. That is the work we do with leadership teams at Radiant Institute, and it is worth a conversation if this is where your organization is stuck.