The BaseResearchInternational Monetary Fund
Bridging Skill Gaps for the Future: New Jobs Creation in the AI Age
An IMF Staff Discussion Note analyses online job postings in six countries. It finds roughly 1 in 10 postings in advanced economies asks for a skill new to the labor market, and such skills go with higher pay; greater demand for new AI skills has not raised overall US employment and goes with lower employment in some exposed occupations.
Why it matters
The same thing, for different readers.
What it means for you
Jobs asking for AI skills pay more, but US regions with more AI-skill demand had lower employment in jobs highly exposed to AI, which the note says challenges young workers.
For lawmakers
The note urges worker mobility, lifelong learning, limits on non-compete agreements and stronger social protection, with priorities depending on a country's Skill Imbalance Index.
For researchers
Shift-share and event-study designs on US commuting zones link posting-based AI-skill demand to American Community Survey employment; online postings overrepresent higher-skill work.
The question it raises
The note closes by saying work provides not only income but also dignity and purpose, and that the role of AI in society needs examining, not only skills policy.
In their own words
Every quotation was found, word for word, in the text we saved from the source. That shows the words are there; it does not make them true, or prove the speaker meant them as quoted.
From the source’s own text.
“In data provided by Lightcast, roughly 1 in 10 job postings requires at least one new skill in advanced economies, and the incidence is about half of that for emerging market economies.”
Found word for word in the source
“Although vacancies demanding new AI-related skills post higher wages, they have so far not boosted overall employment in US local labor markets.”
Found word for word in the source
“Further, the empirical analysis shows that—for occupations that are highly exposed to AI, but with limited scope for complementarity—employment levels are 3.6 percent lower in regions with greater demand for AI-related skills than in other regions five years after the appearance of these skills.”
Found word for word in the source
“These skills boost average wages and employment but deepen polarization, mostly benefitting high- and—through higher consumption of services—low-skilled workers, and potentially contributing to the shrinking of the middle class.”
Found word for word in the source
“Because Lightcast relies on online job postings, the data provide a detailed picture of skill demand in the formal economy but naturally overrepresent professional, technical, and higher-skill occupations.”
Found word for word in the source
How the finding was reached
In our words, from the source. Weigh the finding by how it was made.
- What was done
- Staff analysis of online job postings from Lightcast in six countries (United States, United Kingdom, Germany, Brazil, Denmark, South Africa) to find skills barely present in 2010-11, matched to posted wages; local labor market regressions on US commuting zones (American Community Survey, 2013-23) and 96 German regions (administrative records, 2019-23) with shift-share instruments for new skills; an event study of AI-skill entry on US employment; firm records (Compustat) and nearly 18 million US worker profiles; and skill demand and supply measures for 36 countries.
- What it found
- About 1 in 10 postings in advanced economies lists a new skill, about half that in emerging economies; AI skills were in under 1% of US postings before 2015 and almost 5% by 2025. A new skill goes with 3 to 3.4% higher wage offers in the US and UK. In US local labor markets a 1 point rise in postings with new skills goes with 2.3% higher wages and 1.3% higher employment, gains for high- and low-skilled but not middle-skilled workers. For AI skills alone there is no gain in overall employment: where AI-skill demand is 1 point higher, employment in occupations highly exposed to AI with limited complementarity (about 30% of employment, per the IMF's 2024 work) is 6.3% lower after five years, about 3.6% for the observed rise. AI-skill wage premiums are about 7.5 to 8% in the UK and, in the US, above 8% for developer skills but about 2% for user skills.
- What it does not show
- Measures vacancies and regional employment levels, not layoffs; the note describes the AI result as employment linked to AI-skill demand, and calls its evidence on local adjustment early. Online postings overrepresent formal, urban and higher-skill work, more so in emerging markets, coverage varies by country, and firms that fill jobs internally may be understated. The AI employment result is for US commuting zones; the German analysis covers new skills in general and finds no significant employment effect, which the note says may reflect the short time span and labor market rigidities. The skill indices compare countries with the US as benchmark and cannot quantify mismatches in headcount.
- Status
- A report
The record
- Questions
- What happens to work, income and the economy?
- People
- Florence Jaumotte (Wrote it, International Monetary Fund) · Jaden Kim (Wrote it, International Monetary Fund) · David Koll (Wrote it, International Monetary Fund) · Elmer Li (Wrote it, International Monetary Fund) · Longji Li (Wrote it, International Monetary Fund) · Giovanni Melina (Wrote it, International Monetary Fund) · Alina Song (Wrote it, International Monetary Fund) · Marina Mendes Tavares (Wrote it, International Monetary Fund)
- What we read
- We read the whole source.
- Found by
- A research pass: work, income and the economy, gathered October 13, 2026, via
research-work(tier 1: the source’s own channel or an official record) - Cite this
- International Monetary Fund, “Bridging Skill Gaps for the Future: New Jobs Creation in the AI Age”, January 14, 2026. Entry in the Ctrl AI Base, https://ctrlai.com/base/imf-2026-01-14-bridging-skill-gaps-ai-age
Our notesText: the IMF eLibrary HTML edition of Staff Discussion Note 2026/001 (imf.org and its PDF return 403 to scripted requests; the eLibrary PDF did not download), read in full apart from the annex data tables. Publication date 14 Jan 2026 and DOI 10.5089/9798229028196.006.A001 are from that page; the imf.org address carries 2026/01/09, so the editor may want to check which day the IMF itself gives. The page shows an affiliation only for the first author; all eight are given the IMF because this is an IMF staff note. The page text carries no statement that the note reflects the authors' views and not IMF policy; the research pass reported one in the PDF, which was not read. Quotations use the first author's id. The 6.3% is per percentage point of AI-skill demand; the 3.6% is the predicted effect for the observed rise (0.57 points). The 40%, 60% and 28% exposure shares the note repeats are the IMF's 2024 estimates, not new here. The note cites the NBER ChatGPT paper in this pass (Chatterji and others, 2025) for the speed of AI adoption and the ILO's 2023 exposure study; the editor may want to relate them.