Nearly 90% of Firms See No Measurable Gain From AI. Employees Are Saving a Full Workday a Week Anyway.

Two large, credible surveys landed this year with what looks at first like a contradiction. One, from the National Bureau of Economic Research, found that almost 90% of executives see no measurable company-wide impact from AI after three years of use. The other, from Boston Consulting Group, found that 42% of employees who regularly use AI are personally saving a full workday every week. Read side by side, they aren't actually a contradiction. The second study explains the first.

89%

Executives reporting no measurable productivity impact from AI over the past three years, per NBER

42%

Regular AI users saving a full workday a week, per BCG's survey of nearly 12,000 employees

66%

Of those same employees who get little or no guidance on what to do with the time saved

The NBER Study: Widespread Adoption, Almost No Measured Impact Yet

Economists Nicholas Bloom and Steven Davis, both well known for their work on economic uncertainty, joined a dozen co-authors and central bank researchers at the Federal Reserve Bank of Atlanta, the Bank of England, and the Deutsche Bundesbank to survey nearly 6,000 senior executives across the US, UK, Germany, and Australia. The headline findings, published as NBER Working Paper 34836: 69% of firms actively use AI, with adoption highest in the US at 78% and lowest in Australia at 59%. More than two-thirds of executives personally use AI regularly, but only about 1.5 hours a week on average.

And yet, looking back at the past three years, more than 90% of those same executives report no impact on employment, and 89% report no impact on labor productivity at their firm. Among the minority who do report an effect, the average productivity boost is a modest 0.29%.

The forward-looking numbers are the more interesting part. Those same executives who saw almost nothing happen over the past three years forecast real change over the next three: an average productivity gain of 1.4%, alongside a 0.7% reduction in employment, for a net output gain of roughly 0.8%. US executives are notably more bearish on jobs specifically, predicting a 1.2% employment decline, while their employees expect a smaller productivity gain than executives do, 0.9% versus 2.3%. Executives and the people who work for them are looking at the same technology and drawing different conclusions about where it's headed.

The BCG Study: The Time Savings Are Real. The Redirection Isn't Happening.

BCG's fourth annual Global AI at Work survey, covering close to 12,000 frontline employees, managers, and leaders across more than a dozen markets, offers a plausible explanation for why individual-level time savings aren't showing up in NBER's company-level numbers. Frontline AI adoption has surged to 74% of employees, up 23 percentage points in a single year. Among regular users, 42% report saving a full workday a week, with the heaviest savings concentrated in marketing (60%), IT (53%), and human resources (50%).

Where the Value Is GoingOf the employees saving that time, 66% report getting little or no guidance on what to do with it, and more than half say they aren't redirecting the saved hours into anything their organization would recognize as strategic work. Nearly half now say they spend more time managing and directing AI than doing the underlying work itself. BCG's David Martin, the report's lead author, put the core problem plainly: that saved time can simply leak out of the organization when nobody decides in advance where it should go.

BCG's data also surfaces what it calls a joy paradox: two-thirds of regular AI users report higher job satisfaction, while 41% simultaneously report increased cognitive load. AI is making work both better and harder for the same people at the same time, which is a strange thing for a single tool to do, and a useful reminder that satisfaction surveys and productivity surveys aren't measuring the same thing.

Reading the Two Studies Together

NBER measured the wrong altitude to see what BCG found. Company-level productivity and employment statistics are aggregates. If thousands of individual employees are each saving several hours a week, but that time scatters into busywork, extra meetings, or simply more relaxed pacing rather than concentrating into something a finance department would register as higher output, the company-wide numbers can look completely flat while real, substantial time savings are happening constantly at the individual level. BCG's data backs this mechanism directly: having an explicit strategy for what saved time should be used for lifts AI's measured impact by 25 percentage points, compared with just 5 points from giving people better tools alone. The bottleneck was never really about the technology. It's about whether anyone above the individual employee decided what the saved time was for.

The practical implication for any organization deploying AI: measuring adoption rates or hours saved and stopping there is measuring the wrong thing. The number that matters is what happens to that time next, and right now, at two-thirds of organizations surveyed, the honest answer is nobody has decided.

What This Means for Your Organization

If your organization has rolled out AI tools but hasn't measured whether saved time is being redirected anywhere specific, that gap is exactly what both studies point to as the actual problem, not the tools themselves.
Track where saved time goes, not just whether people are using the tools. BCG's finding that strategic clarity outperforms better tools by a factor of five is the most actionable single number in either study.
Don't assume flat company-level productivity numbers mean AI isn't working. Both studies together suggest it may simply mean nobody has told employees what to do with the time it's freeing up.

Sources: AI Pulse · Big Picture · workplaceai.ai. NBER Working Paper 34836, "Firm Data on AI," by Ivan Yotzov, Jose Maria Barrero, Nicholas Bloom, Philip Bunn, Steven J. Davis, and co-authors, fielded November 2025 through January 2026, published February 2026. BCG's fourth annual Global AI at Work survey, "AI at Work: Why Strategy Matters More Than Tools," published June 2026, as reported by CIO, PR Newswire, and BizTechReports. Every figure above is attributed to its original source; none is a WorkplaceAI study.