---
id: "lib-2026-06-05-ai-and-skills-what-we-know-so-far"
type: "library"
title: "AI and skills: what we know so far"
lang: "EN"
org: "OECD"
source_type: "Report"
date: "2026-06-05"
week: "2026-W23"
url: "https://nosov.ai/library/lib-2026-06-05-ai-and-skills-what-we-know-so-far/"
original_url: "https://www.oecd.org/content/dam/oecd/en/publications/reports/2026/06/ai-and-skills_8dbed5fe/f843b352-en.pdf"
tier: "B"
tier_name: "Useful, limited"
tier_reason: "A nine-page OECD policy brief that restates findings from earlier OECD employer, worker and SME surveys; authoritative source, but no new data and the latest figures date from late 2024."
summary_basis: "full-text"
original_minutes: "12"
tags: ["ai skills", "skill demand", "policy"]
---

# AI and skills: what we know so far

Tier B (Useful, limited): A nine-page OECD policy brief that restates findings from earlier OECD employer, worker and SME surveys; authoritative source, but no new data and the latest figures date from late 2024.

Summary based on the full text.

## In short

An OECD brief dated 5 June 2026 (contact: Stijn Broecke) arguing that skills, more than technology, decide whether economies benefit from AI. Skills shortages hold back adoption, AI raises demand for higher-level skills, and employer training is the main response and seems to work.

## Key numbers

- ~40%: employers in manufacturing and finance not yet using AI who cite lack of skills as the reason
- <1%: share of workers who will need advanced AI-specific skills such as programming or model development
- nearly 40%: SMEs with a skills gap that say generative AI helps compensate for it
- 20% vs 12%: managers in France, Germany, Italy and Spain who say algorithmic management reduces versus increases their need for empathy

## What it reports

- Around 40% of non-adopting employers in manufacturing and finance say skills are the main barrier; more than half of SMEs not using generative AI say the same.
- More than half of AI-adopting employers in those sectors report a greater need for highly educated workers.
- Fewer than 1% of workers need advanced AI skills; most need digital and data skills, plus managerial and human skills.
- Signals of falling demand for some social skills: 20% of European managers say algorithmic management reduces their need for empathy, 12% say it increases it.
- Nearly two in five SMEs faced a worker shortage in the past two years; nearly 40% of those with a skills gap say generative AI helps fill it, a quarter say it helps with a worker shortage. Japan shows the highest shares.
- More than half of AI-using workers received employer-funded training, and trained users report better performance and working conditions more often.
- Policy asks: monitor skill demand, scale access to AI skills for SMEs and at-risk workers, share training costs, tie incentives to AI skills.

## How they got it

Synthesis of OECD surveys: Lane, Williams and Broecke (2023) employer and worker surveys in finance and manufacturing; the 2025 OECD survey of SMEs on generative AI; and the 2025 algorithmic-management employer survey.

## Limits

- No new data; most figures predate widespread generative AI use, and the brief says the newest data are from late 2024.
- Self-reported employer and worker perceptions.
- The training-outcome link is correlational.

## Why it matters for companies

The brief supports budgeting for data-literacy and applied-use training for most staff rather than specialist AI hiring at scale. The empathy finding is a prompt to watch how algorithmic management changes managers' behaviour, not only productivity.

Original: https://www.oecd.org/content/dam/oecd/en/publications/reports/2026/06/ai-and-skills_8dbed5fe/f843b352-en.pdf