Ga direct naar de inhoud
woensdag 18 november 2026Gratis gids
NL
dailyai

AI Jobs Tracker

AI Jobs Tracker · Updated monthly

Loading…

Is AI taking jobs?

Every month we track job ads, layoffs and AI use in six countries. Here’s what’s actually happening, and how exposed your job is.

Try: · · · ·

This month’s biggest stories

New: the agent wave

What happens when AI stops answering and starts doing?

AI agents don’t just answer questions. They carry out whole tasks on their own, like processing a claim, chasing an invoice or booking freight. We re-scored every job for what agents could take on.

How many jobs become highly exposed once agents are counted

Jobs scoring 5 out of 10 or more. The range depends on how capable agents get.

The jobs agents change most

Exposure score today, and the extra once agents are counted. Mostly back-office process work.

Exposure today (ILO)Extra with AI agents
DailyAI Agent Exposure Score (experimental), built by re-scoring the ILO’s 3,265 job tasks with published rules. Agent use: KPMG AI Quarterly Pulse (Q3 2026); Gartner Hype Cycle for AI 2026. See “How we work this out” below.

Fewer ads for jobs AI can do

High-risk jobsMediumLow-risk jobs100 = job ads in Feb 2020
Indeed Hiring Lab job ads; AI risk from the ILO’s 2025 AI exposure scores.

Which countries are hit hardest

How far behind job ads for high-risk jobs are, compared with low-risk jobs.

Indeed Hiring Lab; ILO; Office for National Statistics; Adzuna UK Job Market Report (via Staffing Industry Analysts); DailyAI analysis.

Jobs most exposed to AI

Share of the job AI could do, scored out of 10.

    Jobs least exposed to AI

    These rely on hands-on, physical or in-person work. Common UK and US jobs only.

      ILO–NASK Global AI Exposure Index 2025, 427 occupations. A high score means AI could do many of the tasks. It doesn’t mean the job will disappear.

      Is your field hiring?

      Indeed Hiring Lab job ads, monthly. 100 = Feb 2020. Grey line: all fields in that country.

      Employers want AI skills

      Indeed Hiring Lab AI Tracker. Canada’s jump follows an Ontario rule making employers disclose AI use in hiring.

      Layoffs blamed on AI

      Challenger, Gray & Christmas. US job cuts where the employer named AI as the reason.

      How many companies use AI

      U.S. Census Bureau, Business Trends and Outlook Survey.

      Jonge werknemers

      U.S. Census Bureau, Current Population Survey. Ages 22–25. Earlier = Jan 2021–Oct 2022, latest = Nov 2025–Sep 2026. Highest-risk group vs the rest: +1.0 points (range +0.1 to +1.9).

      For journalists

      Use any chart or figure with credit to DailyAI and a link to this page. Need a custom cut of the data, such as a country, sector or job? Ask us.

      How we work this out

      AI risk. We use the ILO’s 2025 scores for how much of each job’s tasks generative AI could do (shown here out of 10). High risk = the ILO’s two top exposure bands; low risk = jobs the ILO rates not exposed or minimally exposed. For job ads, we group jobs into fields and split them into high, medium and low risk, each covering about a third of US workers.

      AI agents (experimental). For each of the ILO’s 3,265 tasks we ask whether an AI agent, meaning software that plans, uses tools and acts on its own, could carry the task out end to end. Tasks are sorted by their wording and the ILO’s own notes into: physical or on-site work (no change), in-person or authority work (no change), digital process work such as processing, record-keeping, ordering, scheduling, monitoring and coding (large uplift), mixed and other desk work (small uplift). The agent score closes part of the gap between today’s score and 10: half of it for process work in our central case, less in the cautious case and more in the bold one. The rankings barely move between cases.

      Job ads. Indeed Hiring Lab job postings for the US, UK, Germany, France, Canada and Australia, seasonally adjusted, compared with February 2020.

      AI skills. Indeed’s share of ads that mention AI terms such as machine learning or artificial intelligence.

      Layoffs. Challenger, Gray & Christmas announced US job cuts, with the reason the employer gives.

      Young vs older workers. US Census Current Population Survey monthly files, Jan 2015–Sep 2026, ages 22–25 and 30–64, weighted. For each month we compare employment in the fifth of jobs most exposed to AI with the least exposed fifth, for each age group. We fit the 2015–19 trend in that gap and measure how far each later period sits from it. Job codes changed in 2020; we map old codes to new ones using the Census crosswalk (97% of employment matched). Ranges are 95% confidence intervals.

      Company AI use and unemployment. U.S. Census Bureau surveys.

      Important. These figures show what’s happening, not why. Interest rates, the post-pandemic tech hiring boom and each country’s economy also affect hiring. We say so when something else explains a pattern.

      Credits. Exposure scores adapted from Gmyrek, Berg, Kamiński, Konopczyński, Ładna, Nafradi, Rosłaniec, Troszyński (2025), licensed CC BY 4.0. This is an adaptation of an original work by the ILO. Responsibility for the views and opinions expressed in the adaptation rests solely with the author or authors of the adaptation and are not endorsed by the ILO. Job postings and AI Tracker: CC BY 4.0, credit Indeed Hiring Lab. Source: U.S. Census Bureau, Current Population Survey and Business Trends and Outlook Survey.