Businesses are using more AI each month, but the data does not yet show a clear effect on national employment. We judge that from the data below: worker exposure, business adoption, labor market stress, and labor market impact. Click any of them to see how it is built, then scroll for the detail behind each.
Adoption and use
Business adoption
of US businesses report using AI in their business functions in the last two weeks
Share of businesses, by survey wave (every two weeks).
GenAI postings share
of U.S. job postings on Indeed mention generative AI — hiring for building or working with AI
Weekly series, shown as monthly averages.
Worker exposure
of US workers are in occupations where AI is already being used for a meaningful share of the work
Share of US jobs in the most AI-exposed occupations, by quarter.
Automation vs augmentation
of Claude usage is automation, AI doing the task, while 52.6% is augmentation, AI assisting a person
Automation — AI does the taskAugmentation — AI assists
Monthly, Anthropic Economic Index. Covers Claude only.
Early signals
AI-cited layoffs
of total announced U.S. job cuts in August 2026 cited AI or automation as a cause
AI-cited share of announced cuts, monthly (Challenger, Gray & Christmas). August ended a five-month run (March–July) as the most-cited reason.
Entry-level foothold
the gap between recent-graduate (22 to 27) and overall unemployment. Near 1 point is historically normal.
Percentage points, quarterly (NY Fed).
Young developers
employment growth for young developers (under 25) in the most AI-exposed occupation, a young-worker canary
Percent per year, monthly. In-house proxy (computed) from the Stanford Canaries series.
Overall read
Labor market stress
a composite 0 to 100 score tracking unemployment, hiring and layoffs, nationally
A custom metric defined by aidisruption.fyi.
Labor market impact
Inconclusive
It is not yet clear whether AI is affecting the national labor market.
Jobs in AI-exposed industriesJobs in all other industries
Quarterly, indexed to 100 in 2015. Differences so far are within normal variation.
What businesses say they use, and what they are hiring for
Share of US businesses using AI, by survey wave
Estimated share90% interval
The wording changed partway through, so compare before and after with care. Cross-check: a separate Stanford survey moves in the same direction.
The demand side: job postings that mention AI
A second angle on adoption. Surveys ask firms what they use; postings show what they are hiring for.
Share of postings mentioning generative AI. Weekly series, shown as monthly averages. Available by metro, so a metro profile can carry it later.
Break it down by
Who and where is AI showing up?
Two views of the same question. Pick a metro to open its profile, or scan the occupations that carry the most AI use.
AI in your community
Loading map data…
Exposure vs national average:
Highest exposure
Exposure compares a metro's job mix with the national average. It is about the mix of jobs, not about layoffs.
Occupations carrying the most AI use
Occupation
Workers
Used today
Could be done
Automated vs augmented
AutomationAugmentation
Ranked by workers weighted by observed AI use. Usage data covers Claude only.
Are there early warning signs?
Leading indicators that tend to move before unemployment does. Open any card for the chart, the data and how it is measured.
Latest values, changes and as-of dates are from the live site. Open any card for the chart, the data and how it is measured.
Sources and methods
Anthropic Economic Index, O*NET, BLS OEWS, JOLTS and CPS, Census Business Trends and Outlook Survey, Quarterly Workforce Indicators, WARN notices, Challenger, Gray & Christmas, Indeed Hiring Lab. Each source and each calculation is documented on the methods page, with downloads.