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.
Editorial responsibility: The Impact of AI Editorial Desk · Report a factual concern
Reads the full article in a natural voice. First play may take a moment to prepare.
Research topic
Which AI deployments deliver independently measured gains after oversight, integration and error costs are included, and how are the gains shared with workers?
At a glance
- 1McKinsey's respondents report much wider personal productivity benefits than company-wide earnings effects: 80% describe a personal gain, while 37% attribute any positive EBIT effect to AI.
- 2The survey records 44% reporting enterprise-scale AI adoption, but only 6% fit McKinsey's own high-performer definition; operating cost constrained use for about one in five.
- 3Expected job cuts should not be read as observed job losses: 39% foresee reductions next year, while 14% say AI contributed to an overall reduction in the past year.
Living evidence record
Impact record IAI-1WNZ43I
Evidence stage
Studied
Confidence
Supported
Reporting basis
Multi-source analysis
Independent support
Not yet
Record status
Monitoring
Last checked
28 September 2026
Source trail
2 direct sources 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.
What McKinsey actually measured
McKinsey's State of AI survey, published on 25 August, asked 1,719 online respondents in 97 nations about their organisations' AI use between 4 May and 8 June 2026. It weighted responses by each nation’s contribution to global GDP; 36% of respondents worked at organisations with annual revenue above $1 billion. These details matter because the results are answers from a broad group of professionals, not audited balance sheets or a census of every business. The newsroom is examining an August source on 28 September, so this is a new analysis of an earlier report rather than a newly released survey.
Eighty percent of respondents say AI improved their own productivity, and half say it helped them make better decisions. Yet 37% attribute at least some positive effect on their organisation's earnings before interest and taxes to AI, essentially unchanged from last year. Six percent meet the report's stricter high-performer threshold: at least 5% EBIT impact attributed to AI plus a description of significant value. These are separate questions and denominators. A worker can save time without that gain appearing in firm-wide profit, especially when tools, review, integration and redesign also cost money.[1]
Adoption, costs and the workforce
Nearly nine in ten McKinsey respondents report regular AI use in at least one business function; 44% say their organisations are scaling it across the enterprise, up from 38% a year earlier. Among respondents at companies with at least $1 billion in revenue, 40% report scaling agents in a function, versus 22% at smaller organisations. About one in five say AI operating costs have constrained use. Nearly one-third report deciding against buying at least one software product or feature because their organisation could build it with coding agents. These figures describe respondent reports about different kinds of use and purchasing decisions. They do not establish that every agent deployment is profitable.
Stanford HAI's April 2026 AI Index provides a useful wider frame: it describes rising organisational adoption but says agent deployment remained in single digits across nearly all business functions in its underlying evidence. Its measures, collection windows and definitions differ from McKinsey's August survey, so the percentages cannot be subtracted from one another or treated as replication. Together, the sources suggest a distinction worth investigating in a company: access to an AI tool, scaled use in a department, and sustained value after all costs and errors are counted are three different milestones.[1][2]
What would prove durable value
The employment numbers need equal care. Thirty-nine percent of respondents expect AI to reduce total head count in the next year, compared with 43% expecting little or no change. McKinsey says 14% of respondents at AI-using organisations report that AI contributed to an overall decline in the past year. Last year's 32% expectation of reductions was therefore much larger than this year's reported outcome. Neither number isolates AI from hiring freezes, demand, interest rates or other business changes. The survey records perceptions and plans; it cannot demonstrate how many jobs AI caused to disappear.
The report associates stronger financial returns with redesigning workflows and leadership practices, but this is an observational association among survey respondents, not a randomised test of those practices. For readers evaluating an investment, the better question is what changes in a specific process: task time, quality, error correction, staff supervision, customer outcome and full operating cost, measured against a credible baseline. Report results separately for large and small organisations and for workers affected by the redesign. A later independent assessment of real accounts and employment records could test whether reported personal gains compound into lasting value and whether the burden of change is fairly shared.[1][2]
What this means for people
- Workers may gain time on routine tasks yet face more review work, changed skills and uncertain hiring; expected workforce cuts are not observed job losses.
- Smaller organisations may face a different path to returns because their capacity for integration, supervision and risk controls differs from that of large enterprises.
Global context
McKinsey's online sample spans 97 nations but is weighted by GDP contribution, and 36% of respondents work at organisations above $1 billion in revenue. Stanford's AI Index aggregates other underlying datasets. Neither source measures the typical result for every country, sector or small employer.
What the evidence does not yet show
- McKinsey's figures rely on respondents' self-reports and attribution of outcomes to AI; the survey does not independently audit earnings or establish causation.
- Stanford's April synthesis uses other datasets and definitions, so its adoption figures are context rather than independent validation of McKinsey's percentages.
- A management consultancy that sells AI advice has a commercial interest in the subject. Its method and findings should be examined on their own terms and against independent evidence.
What to watch next
- Repeated, comparable measurement of task quality, full operating cost and earnings after rollout, including oversight and failures.
- Observed hiring and wage data by age, role, country and sector, compared with the survey's workforce expectations.
Evidence trail
Sources used for this report
Links checked 28 September 2026
This report is labelled multi-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.
Continue the story
Related reporting
Finance & Business
AI infrastructure borrowing forecast at $420bn in 2027 as bond buyers demand more
Goldman Sachs data cited by Reuters projects record gross hyperscaler issuance next year. Investors are asking whether data-centre returns justify the scale and concentration of borrowing.
5 min · 1 source
Finance & Business
Who receives AI productivity gains may matter for inflation, central banker argues
ECB Governing Council member Fabio Panetta said central banks need to understand how AI-driven gains are distributed because the split between wages, profits and prices will shape demand and inflation.
4 min · 1 source
Finance & Business
BIS says the AI investment boom can blur economic signals for central banks
A BIS bulletin argues that AI affects demand, supply, trade and asset prices at different speeds, complicating the task of reading inflation and underlying economic capacity.
4 min · 1 source
Reader discussion
Add evidence, experience or a question
No account is required. Reader notes are published after a brief civility, relevance and safety check; disagreement is welcome.
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.