~130 occupation groups · 36 countries · 199.6M jobs
~130 occupation groups · area = employment · color = AI exposure score (0–10)
Read the analysis →Each of the 125 European occupation groups was scored by Claude Sonnet 4 on two dimensions:
Technical Exposure (0-10): How much could AI reshape this occupation's core tasks? Based on ESCO occupation descriptions covering skills, tasks, and work context for 3,043 detailed occupations aggregated into 125 ISCO groups.
Regulated Exposure (0-10): Same assessment but accounting for EU AI Act obligations, GDPR constraints, works council requirements (DE: BetrVG §87, AT: ArbVG §96a), and sector-specific regulation. The average EU regulatory delta is 1.2 points — meaning regulation reduces effective AI exposure.
UK Regulated Exposure: Separate scoring reflecting the UK's lighter regulatory framework (avg friction: 0.5 vs EU's 1.2).
Additional layers use official statistical sources: Cedefop Skills Forecast 2025 (employment growth to 2035), Eurostat Labour Force Survey (education levels, pay), and triangulated AI adoption data from Anthropic, Microsoft, and OpenAI research.
Rate each occupation group's overall AI Exposure on a scale from 0 to 10.
AI Exposure measures how much AI will reshape this occupation. Consider both
direct effects (AI automating tasks currently done by humans) and indirect
effects (AI making each worker so productive that fewer are needed).
A key signal is whether the work product is fundamentally digital. If the
occupation can be performed entirely on a computer — writing, coding,
analysing, communicating — then AI exposure is inherently high (7+).
Conversely, occupations requiring physical presence, manual skill, or
real-time human interaction have a natural barrier.
Calibration anchors:
0–1 Minimal change expected. Almost entirely physical or hands-on work.
Examples: Roofers, building caretakers, agricultural labourers.
2–3 Low change expected. Mostly physical or interpersonal. AI assists
with peripheral tasks; core work remains human-driven.
Examples: Electricians, plumbers, firefighters, dental assistants.
4–5 Moderate change expected. Mix of physical and knowledge work. AI
assists with information-processing parts; roles evolving.
Examples: Registered nurses, police officers, veterinarians.
6–7 Significant evolution. Predominantly knowledge work with some need
for human judgment or physical presence. AI tools reshaping workflow.
Examples: Teaching professionals, managers, accountants, journalists.
8–9 Rapid evolution. Almost entirely computer-based. Core tasks —
writing, coding, analysing, designing — are in domains where AI
is creating new ways of working.
Examples: Software developers, legal associates, data analysts,
translators, graphic designers.
10 Maximum evolution. Routine information processing, fully digital.
Roles transforming into oversight, quality assurance, and exception handling.
Examples: Data entry clerks, medical transcriptionists.
Starting from the technical exposure score, adjust for regulatory friction that slows, constrains, or redirects AI adoption in this occupation group. Consider these regulatory dimensions: EU AI Act (Regulation 2024/1689): - Is this group a high-risk AI deployer under Annex III? - Does Art 26(7) require worker information obligations? - Does Art 14 require human oversight mechanisms? Data Protection: - Does GDPR Art 22 restrict automated individual decision-making? - Does GDPR Art 35 require data protection impact assessments? Employment Law (country-specific): - Germany: Does BetrVG §87 give works councils co-determination rights over AI tool introduction? - Austria: Does ArbVG §96a require works council consent for employee monitoring systems? - Switzerland: Do FADP Art 21, OR Art 328b, or ArGV3 Art 26 apply? Sector Regulation: - Platform Work Directive applicability - Pay Transparency Directive implications - Sector-specific licensing or professional standards The regulated score should be ≤ the technical score. The difference (technical minus regulated) represents the regulatory buffer — the friction that buys workers and institutions transition time. Average EU regulatory delta: 1.2 points. Average UK regulatory delta: 0.5 points.