Gregory Thwaites and his colleagues have just completed the largest cross-country survey of how firms are actually using artificial intelligence—and the answer is: far more widely than many expected. The study, based on candid conversations with over 5,000 CEOs, CFOs and senior executives across the United States, United Kingdom, Germany and Australia, reveals that roughly seven out of every ten firms are already deploying some form of AI technology in their operations today.
Yet here lies the paradox at the heart of the artificial intelligence moment: despite this remarkably high adoption rate, most businesses report seeing little measurable impact on employment or productivity yet. A gap exists between the hype surrounding AI and the on-the-ground reality in boardrooms across four of the world's largest advanced economies. For workers and policymakers anxious about disruption, this might sound reassuring. But the survey tells a more complex story.
The research, conducted through harmonized survey questions administered across economy-wide business surveys in each country, examined not just whether firms have adopted AI, but how they're using it and what they expect it to do. The findings show a striking divergence between what executives see happening now and what they believe will happen in the next three years. While current measurable effects remain modest, business leaders anticipate significantly larger impacts ahead—including notable productivity gains. The same executives, however, also expect reductions in employment growth, suggesting they believe AI will help firms do more with fewer workers.
This creates a crucial tension in the data: the gap between executive expectations and employee expectations regarding AI's impact on the labor market. The study explicitly highlights these differences, raising fundamental questions that go beyond spreadsheets and deployment timelines. What does it mean when the people making decisions about AI adoption expect different outcomes than the people whose jobs may be affected by those decisions? The researchers suggest this divergence has real implications for how societies should think about technological change, firm dynamics, and the future of work itself.
The paper, published as an NBER Working Paper and accompanied by an interactive data dashboard, represents the most comprehensive snapshot yet of AI adoption across major English-speaking economies. It draws on contributions from over a dozen researchers, including economists from the Resolution Foundation, the Bank of England, and international institutions. Gregory Thwaites, a Research Associate at the Resolution Foundation and Associate Professor at the University of Economics at Nottingham, led the effort alongside co-authors including Nicholas Bloom and Steven J. Davis, scholars known for their rigorous work on firm dynamics and economic measurement.
What makes this research valuable is its refusal to settle for easy answers. The study doesn't simply ask "Is AI being adopted?" or "Will AI create or destroy jobs?"—questions that often yield polarized responses depending on who's answering. Instead, it asks what is actually happening in firms right now, what executives genuinely believe will happen, and where the gaps in perception lie. As artificial intelligence continues to reshape work across sectors, evidence grounded in what thousands of business leaders actually see in their firms may matter more than speculation. The dashboard released alongside the paper invites ongoing exploration of the data, suggesting this conversation is far from over.