The AI productivity boom that hasn’t happened yet

A newly released National Bureau of Economic Research(NBER) working paper analysed responses from nearly 6,000 CEOs, CFOs, and senior executives across four major economies.

Here's what they found:
→ Over 80% of firms report no measurable impact of AI on productivity
→ 70% of those firms are already actively using AI
→ Top executives using AI average just 1.5 hours a week, and 1 in 4 report no AI use at all

Let that last one land. These are the people signing off on the AI budgets.

The PwC 2026 Global CEO Survey pointed in the same direction. 4,454 CEOs across 95 countries. Only 12% reported AI delivering both cost and revenue benefits. 56% said they had seen no significant financial impact.

And yet, the same NBER study found executives predict AI will meaningfully boost productivity over the next three years. They believe it's coming. They just can't find it yet.

Economists have a name for this pattern. In the 1980s, Nobel laureate Robert Solow observed that "you can see the computer age everywhere, except in the productivity statistics." Almost forty years later, they're saying the same thing about AI.

One way to interpret this is simple.

AI was dropped into organisations built for a different era. Roles, workflows, and decision rights are all designed for slower, more manual conditions. When AI enters an organisation, it tends to illuminate the assumptions the structure was built on, and where those assumptions no longer hold.

The technology works. The structure around it struggles.

The wreck was already there.
AI just gave us the visibility to see it clearly.

The organisations beginning to close this gap tend to look in a different direction. They examine the structure underneath the work, how decisions actually move, where capability truly sits, and what a role becomes once AI takes over the repeatable parts.

That's harder than buying another subscription.

It's also where the gains are hiding.

At iForce Connect.uk, we help organisations understand how work, skills, and decision-making need to evolve together, not just which tools to deploy.