Surveyed leaders report limited efficiency improvement, modest output expectations, minimal job cuts, and underuse of generative tools within organizations.
According to a recent survey of nearly 6,000 executives in the United States, United Kingdom, Germany and Australia by the National Bureau of Economic Research (NBER) has had more than 80% of respondents from firms adopting AI citing that the productivity boom promised by the technology was overrated, even as managements continue to pour money and optimism into the technology.
The responses were elicited from CEOs, CFOs and other senior leaders at organizations of varying sizes, most of which have already begun experimenting with AI tools.
Despite the lack of measurable gains so far, adoption was widespread: about 70% of respondents cited they were using some form of AI. Yet, more than four in five of those respondents reported no effect on productivity or headcount, underscoring a growing gap between investment and results.
However, respondents remained bullish about what AI could deliver over the next few years. On average, respondents expected AI to raise productivity by 1.4%, increase output by 0.8% and cut employment by 0.7% over a three‑year horizon. Many of them barely used the tools themselves: around one‑third of respondents cited using AI at work, but for 1.5 hours a week on average, while a quarter reported no use at all.
The findings feed into a broader debate over whether today’s wave of generative AI (GenAI) will ultimately mirror past technology revolutions, which often took years to show up in productivity statistics. Recent analysis for the Organisation for Economic Co‑operation and Development (OECD) argues that GenAI can deliver sizeable efficiency gains in tasks such as writing, editing and coding, but only when organizations integrate the tools effectively into workflows and build human skills around them.
OECD researchers have noted that firms and entrepreneurs that carefully align AI with business processes and decision‑making tend to see the largest benefits, while others risk adding complexity or even setbacks.
Combined with the NBER survey’s finding that most surveyed firms reported minimal productivity boosts so far, the emerging picture suggests that the constraint is less about access to AI than about how seriously leaders redesign work around it.
