Skip to content
Start here

Start here

A short guide to what this site is, the questions it can help you answer, and where to look first. If you have never visited before, this page is the fastest way in.

What this site is

aidisruption.fyi tracks public labor-market data and published AI-exposure research to give an evidence-based view of how AI is affecting the US workforce. It is a monitoring tool, not a commentary site: it shows where and how quickly labor-market patterns are changing, and which careers and metropolitan areas are most affected. It does not predict, and it does not editorialize.

The site aggregates a range of public data sources into three composite indices — labor-market stress, structural AI exposure, and community adaptive capacity. Every number is built from named public sources, and every measure's limits are stated next to it. It is maintained independently by Dan Rohan; About This Site explains who runs it and why.

Four questions this site can help you answer

Is AI adoption real?

Business adoption comes from the Census Bureau's Business Trends and Outlook Survey, which asks a sample of US businesses whether they used AI in the last two weeks, wave after wave. Worker exposure comes from published usage data showing how much AI is actually used on real work tasks. Both series are shown live on the National Pulse, with their sources disclosed. One honest caveat: a business saying it uses AI says nothing about how much of its work is affected.

See the live adoption and exposure measures →

Are businesses building with it?

Alongside the adoption survey, the site tracks the share of US job postings that mention generative AI — hiring for building or working with AI — and which industries are doing the hiring. That is a hiring-demand signal, not a headcount: a posting mentions the technology long before anyone can know what actually changed on the ground.

How adoption and postings are measured →

Are layoffs being attributed to it?

No — and the site is careful not to pretend otherwise. Layoff notices and the labor-market stress measures show what is happening to employment, but none of them shows that AI caused it. The site's own labor-market impact tests came back Inconclusive: the recorded verdict is that the current research design cannot even isolate the comparison, so the impact card says so plainly instead of showing a number. Where employers themselves cite AI in layoff reports, the site tracks that as a reported signal — reported is not proven.

See stress and the impact card →  ·  The tests behind it →

What does it mean for a specific job or place?

Two tools answer the personal versions of the question. The Occupations page scores an occupation for AI exposure, shows task by task what AI is used for today and what it could do, and shows hiring and reported layoffs for that job. The Community page does the same for a county, city, or ZIP code: how exposed the local mix of jobs is, which industries carry the most AI use, and recent layoff notices nearby. Both are comparisons, not verdicts — a high score means more AI use or more stress, not that the job or place is going away.

Look up an occupation →  ·  Look up a community →

What this data cannot tell you

Nothing here establishes causation. No single layoff, hiring change, or shift in unemployment can be attributed to AI alone — the economy has too many simultaneous drivers. The scores show patterns that are consistent with AI-driven displacement, not proof of it.
Exposure is not job loss. A highly exposed occupation is one where a lot of AI-assisted work happens. That says nothing by itself about whether employment there will fall.
The methodology is provisional. The way every score is computed is being overhauled, and until that work lands and is verified, the numbers on the site should be treated as provisional. See how scores are computed.
No predictions. No forecasts are published. The scenario tracker watches published indicators, but scenario probabilities stay hidden until a calibrated, backtested model exists — what you see is the evidence, not a verdict.

A path through the site

If you want one place to begin, go in this order:

Start with the National Pulse
The national picture: worker exposure, business adoption, labor-market stress, and the labor-market impact card. It is fed from live data, and every measure links to how it is built.

Open the National Pulse →

Look up your occupation
Search the Occupations page for a job title: how exposed it is, the task-by-task breakdown of what AI does today, and what hiring looks like.

Look up an occupation →

Check your community
Search the Community page for a county, city, or ZIP code: how exposed local jobs are, how much stress the local market is under, and recent layoff notices near you.

Check your community →

When you want the depth behind any number, How scores are computed has one section per score, including what each one cannot tell you. And if something looks wrong, the corrections policy explains how errors get fixed in the open.