Policy
Policy briefs, technical reports and policy-facing analysis.
European Commission
AI and Skills
with Alice Bertoletti, Alessia De Santo and Wouter Vergote
Publications Office of the European Union, 2026 · JRC148616
TL;DR Abstract JRC repository
Exposure to artificial intelligence has risen across the occupational spectrum, but actual use lags behind: in 2024-2025, 30% of EU workers reported using AI at work, and by 2025 20% of enterprises had adopted AI technologies. Adoption is uneven, ranging from close to 40% of white-collar workers to 6% in elementary occupations, and concentrated in large firms, digitally intensive sectors and Northern Member States. Exposure is highest in high-skilled, office-based work, but it does not equal replacement: whether AI augments or displaces workers depends on how central judgement, responsibility and human interaction remain to the job. So far, employment and vacancy data show no clear sign of labour-market disruption, which makes it difficult to design targeted reskilling measures and calls for more granular data and monitoring. On the supply side, AI-related education and training in EU higher education is anchored in master’s degrees, while the short, flexible formats most needed by the already-employed workforce remain underdeveloped. AI literacy should go hand in hand with transversal and human-centred skills, teachers need strong policy support to adapt, and AI in vocational education and training should be approached as a system transformation, building on proven initiatives such as those under the Pact for Skills.
AI exposure is highest in high-skilled office work, but use lags well behind: 30% of EU workers and 20% of firms use AI, with no clear sign yet of labour-market disruption. On the supply side, the main gap is short, flexible training for people already in work.
European Skills Intelligence Observatory: Mapping of Skills Intelligence in the EU (Technical Report)
with Wouter Vergote
Publications Office of the European Union, 2026 · JRC147896
TL;DR Abstract JRC repository
Qualification mismatch and skills mismatch overlap only modestly, and measured incidence depends on the instrument, so household, employer and vacancy sources complement rather than replace one another. The report scores candidate indicators on seven criteria for use in the ESIO dashboard.
This report maps the data sources and indicators available for measuring skills demand, supply and mismatch across the European Union. It assesses which of them can support regular monitoring by the European Skills Intelligence Observatory (ESIO). Each instrument is profiled on a common template covering description, methodology, published indicators, granularity, coverage, update frequency, microdata access, strengths and limitations. The mapping spans Eurostat surveys and online job advertisement data, the skills intelligence system of Cedefop, shortage and surplus reporting by the European Labour Authority, Eurofound surveys, sectoral partnerships, the OECD adult skills programmes, and private vacancy analytics. Three results follow. First, qualification mismatch, measured through credentials, and skills mismatch, measured through assessed or reported competencies, overlap only modestly. Hence, household, employer and vacancy sources are complementary rather than substitutable. Second, measured incidence depends on the instrument, because wording and thresholds differ even among sources asking about the same concept. Third, many of the instruments are also interdependent, each inheriting the constraints of the sources it draws on. The report converts this mapping into a scoring framework in which candidate indicators are rated on seven criteria. Scoring is comparative rather than threshold based, and records fitness for the ESIO dashboard use rather than quality
European Skills Intelligence Observatory: First brief to the European Skills High Level Board
with Montserrat López Cobo and Wouter Vergote
Publications Office of the European Union, 2026 · JRC147054
TL;DR Abstract Brief JRC repository
This policy brief by the European Skills Intelligence Observatory (ESIO) highlights critical skills challenges facing the EU. Different sources of evidence point towards skills mismatches that are particularly acute in ICT, engineering, health and care, and skilled trades. These shortages diverge across countries, and tackling them requires different levels of education and training. Evidence shows that the supply side is not keeping up due to issues such as the decline in basic skills, teacher shortages, and a large untapped potential of working-age adults who remain outside the labour force or are unemployed or underemployed. In particular, STEM enrolment, which is instrumental for EU competitiveness, is not increasing in line with demand. This gap will widen unless the talent pipeline is broadened, especially for women. The brief also emphasises that AI adoption is reshaping demand for digital and transversal skills and tends to widen the skills gap, calling for effective lifelong learning policies in addition to adapted formal training programmes. The brief underscores the need for improved skills intelligence to address mismatches and future-proof Europe’s workforce.
Skills shortages in the EU are most acute in ICT, engineering, health and care, and skilled trades, and supply is not keeping pace. Closing the gap requires a broader STEM pipeline, especially for women, and lifelong learning that keeps up with AI adoption.
Other policy reports
What has been happening to firm creation by UK region and sector?
with Anthony Savagar
Economics Observatory, 2021
Business creation fell across the whole of the UK during the first lockdown but boomed during the second and third. Certain regions and sectors have recovered faster than others, notably London, and wholesale and retail trade.