CFOs expect less than 0.4% AI job loss in 2026—but are they measuring or forecasting?
A 734-executive study found broad but shallow AI adoption and modest expected aggregate employment effects. The figures come from company leaders' attribution and forecasts, not observed economy-wide job losses caused by AI.
The short version
- What happened
- Two surveys near the end of 2025 yielded 734 senior-finance-executive responses across 48 US states, major industries and firm sizes.
- Who it affects
- Workers in routine clerical roles may face task redesign even if total employment changes remain modest. Managers should separate purchased subscriptions from measured gains and should track who receives training, whose tasks disappear and whether workload or pay changes.
- What remains unproven
- Survey responses mix reported past effects with expectations and do not identify a causal counterfactual.
Keep in mind
Survey responses mix reported past effects with expectations and do not identify a causal counterfactual.
The sample consists of senior financial executives; worker accounts, payroll records and audited productivity were not the primary evidence.
Weights map the sample to broad sectors and firm sizes but cannot remove nonresponse, interpretation or optimism bias.
The study focuses on incumbent firms and largely excludes new AI-intensive firms and upstream infrastructure investment.
Source analysis · 1 linked source. Primary analysed source date: 30 October 2026.
Keep following the evidence
Join The Impact Brief founding list for selected AI reports, their effect on people and the evidence limits. Read a sample first.
Read the full report · 3 min The evidence, methods, implications and limitations
Adoption was widespread, but most spending remained shallow
Researchers from the Federal Reserve Banks of Atlanta and Richmond, Duke University and the NBER combined the CFO Survey with responses from financial executives affiliated with FEI, Nasdaq and Duke. The 734-company sample covered 48 states and every major nonfarm nongovernment industry. Representativeness weights aligned sectors and firm-size groups with the US economy; employment weights gave larger employers more influence when estimating aggregate effects.
Most AI spending was operational—subscriptions, services and training—rather than hardware or internal development. Median planned 2026 spending was $12,500 at firms below 500 employees and $300,000 at larger firms. Executives associated AI with faster decisions, higher-value work and product development. Yet the paper describes adoption as broad but shallow: paying for tools is not the same as reorganising production around them.[1]
The headline job estimate is an expectation, not a count of observed layoffs
After weighting responses to sector employment, the authors estimated 501,836 fewer jobs attributed to AI in 2026—less than 0.4% of the workforce. Half of firms explicitly said AI would replace no role, while 6% were unsure. Expected losses rose with firm size and clustered in finance and high-skill services. By 2028, respondents anticipated the share of routine clerical work falling by more than two percentage points, partly offset by technical roles, especially at smaller firms.
Those numbers should not be read as a verified forecast. Executives may interpret questions differently, be optimistic about productivity, or attribute changes to AI that have several causes. Revenue-per-worker gains can also reflect prices, demand or mark-ups rather than technical efficiency. For workers, the practical signal is task composition: routine administration may face more pressure than total headcount. Stronger evidence would compare forecasts with payroll and task data over time, include failed adopters and new firms, and test whether reported productivity appears in audited output.[1]
The distinction that matters
Three layers of evidence
The paper combines current adoption, company attribution and future expectations.
Observed input
Survey 734 financial executives
Broad adoption, mostly subscription and service spending
Attributed result
Ask firms what 2025 changes came from AI
Reported productivity gains, with perception and attribution limits
Forecast
Ask about 2026 and 2028
Modest net employment decline and larger clerical-task shifts expected
Watch the background, then check the evidence
Illustrative photo · not a video frame
UNESCO · 11 min 27 sec · Background explainer
UNESCO Talks: Interview with 2024 Nobel laureate Professor Daron Acemoglu
YouTube loads only when you choose; playback starts when you press play.
Economist Daron Acemoglu discusses how choices about AI affect workers, inequality and society. Pair the interview with our comparison of task exposure, hiring and skills evidence.
Evidence check: An expert interview from 2024, offering a perspective on policy and work rather than a measured forecast of job losses.
Published 2024-10-14. Source checked 14 October 2026. Check captions on the publisher’s video page.
Read the AI and jobs guidePhoto: Dylan Gillis / Unsplash · Licence
Continue the story
Read next
Continue exploring productivity: compare the finding and its evidence limits.
McKinsey's 2026 AI survey: individual gains are clearer than company-wide returns
In a 1,719-person global survey, more respondents report AI helping their own productivity than report a measurable effect on company earnings. The gap deserves scrutiny, not a promise of inevitable returns.
Read the next reportExplore the full source and evidence record
What this means for people
- Workers in routine clerical roles may face task redesign even if total employment changes remain modest. Managers should separate purchased subscriptions from measured gains and should track who receives training, whose tasks disappear and whether workload or pay changes.
Global context
The study is US-specific, but its distinction between shallow adoption and organisational change travels. International comparisons need local labour institutions, firm-size distributions and adoption costs rather than importing the estimated US headcount effect.
What the evidence does not yet show
- Survey responses mix reported past effects with expectations and do not identify a causal counterfactual.
- The sample consists of senior financial executives; worker accounts, payroll records and audited productivity were not the primary evidence.
- Weights map the sample to broad sectors and firm sizes but cannot remove nonresponse, interpretation or optimism bias.
- The study focuses on incumbent firms and largely excludes new AI-intensive firms and upstream infrastructure investment.
What to watch next
- Forecasts compared with payroll data
- Routine-task changes within jobs
- Training and wage effects
- Audited output per worker
- Small-firm technical hiring
- Evidence from other economies
Living evidence record
Impact record IAI-1HPO5EF
Evidence stage
Studied
Confidence
Supported
Reporting basis
Source analysis
Independent or research support
Present
Record status
Monitoring
Last checked
31 October 2026
Source trail
1 direct source across 1 source type.
People impact
Documented in this record.
Uncertainty
Limits and next checks are explicit.
Stages describe the evidence available—not whether a technology is good or bad. See the public method.
Single-source reporting disclosure
This record analyses one direct source. It can establish what Journal of Finance: Insights and Perspectives published or reported, but it is not independent corroboration of every performance claim or predicted outcome. The confidence label will change only when broader evidence is added.
Evidence trail
Sources used for this report
Links checked 31 October 2026
Research question, topics and regions
Topics: Workforce · Productivity · CFO survey · Employment · Business adoption · Forecasts
Regions: North America · United States
This report is labelled source analysis. We summarise and analyse source material in our own words; company statements remain attributed claims until independently supported. Translated summaries preserve the meaning of the original source and link back to it. Read our editorial standards.
Reader commentary
What does this mean where you are?
Tell us what you have seen at work, in your community or in everyday life. A question is welcome too. No account is needed; we review notes before publishing them.
Explore commentary across the portal →Published reader notes
0No published reader notes yet. You can start the evidence-led discussion above.
Prefer a private correction or response? Contact the newsroom.