About
We count AI jobs, and we show our working
What we publish
- Level 1
The tracker
We poll the job feeds employers publish themselves, every day, and count the AI roles open across the set we track. That count is ours. Nobody else publishes it.
See the tracker - Level 2
The digest and the monthly report
Each morning we read the day's AI-and-employment news against our own count and write up what the two say together. Once a month we publish the same thing over a longer window.
Read the digest - Level 3
The news we track
The reporting the digest is written from. It is other outlets' work, every entry links back to them, and it is deliberately kept out of search results because it is source material rather than something we wrote.
Browse the database
What we do not claim
- That the headline figure is a raw count. It is calibrated to the World Economic Forum's Future of Jobs Report 2025, and the arithmetic is written out in full on the methodology page.
- That it is a census of the labour market. It covers the employers in our tracked set, scaled up. It is not a survey.
- That we can predict what happens next. We publish what we counted and what was reported, with the date we counted it.
The rules we work to
We publish aggregate counts, not employers' posting text. We honour robots.txt and the rate limits of every feed we read, and we exclude sources whose terms prohibit polling. Where a figure is modelled rather than observed, we label it modelled on the page it appears.
Every data surface on the site carries a stamp saying what the figure is as of and where it came from, and that stamp links to the full method.
Who publishes this
AI Jobs Report is published by Straits AI Group.
Write to contact@aijobs.report for anything about the data, and to media@aijobs.report if you are citing a figure and want it checked before you publish. We would rather answer a question than see a number misquoted.
Corrections. If a figure here is wrong we correct it and say what changed. Counts are restated rather than silently edited, and the methodology page records when a definition changes — the panel has grown repeatedly, most visibly from 49 employers to 119 on 15 August 2026, and stands at 200 today, and every like-for-like comparison on the site is computed to be unaffected by that. Tell us what looks wrong and we will either fix it or show you the working.
The data is yours
The counts are published under CC BY 4.0. Use them, chart them, quote them in your own reporting. We ask only that you link back so a reader can check the method for themselves.