Living dashboard
AI Labour Impact Observatory
AI's impact on employment
A living dashboard tracking the impact of artificial intelligence on employment, combining global sources, market signals and our own interpretation.
This observatory does not measure jobs created or destroyed alone. It measures the pressure of labour transformation and the capacity to absorb it.
- IILE-IA v0.2
- Curated monthly
- Weekly source watcher
- Experimental
Evidence × Impact
Evidence × Impact
Editorial reading of each verified claim against two axes: evidence strength and expected labour impact.
Expected labour impact ↑
Market expectations (survey, high impact)
Employer surveys and market forecasts.
World Economic Forum — Future of Jobs Report 2025
Global employer survey
World Economic Forum — Future of Jobs Report 2025
Global employer survey
Theoretical exposure (strong evidence)
Academic / multilateral methodologies with high labour impact.
IMF — Gen-AI: Artificial Intelligence and the Future of Work
Macroeconomic + labour analysis
IMF — Gen-AI: Artificial Intelligence and the Future of Work
Macroeconomic + labour analysis
ILO — Generative AI and Jobs: A Refined Global Index of Occupational Exposure
Occupational exposure research
ILO — Generative AI and jobs: A 2025 update
Occupational exposure research
Weak signals
Priority sources without an extracted claim yet.
Stanford HAI — AI Index 2025 Economy chapter
Labour-market data analysis
Real adoption (strong evidence, observed impact still bounded)
Adoption / deployment data with bounded effects so far.
Pending: no observed real-adoption claim yet.
Survey / expectation
Traced methodology
Evidence strength →
The matrix is not a statistical scatter plot. A cell position is an editorial zone, not a quantitative coordinate.
Evidence vs hype
Three angles so we don't confuse what AI could do with what AI is doing.
Theoretical exposure
- What it measures
- Share of tasks that AI could automate or augment, according to models.
- Examples
- Stanford HAI, OECD, MIT, Brookings, NBER papers.
- Why it matters
- Marks the ceiling of change — the maximum room for movement if the tech were fully deployed.
- If misread
- Mistaking it for reality. Theoretical exposure tends to overstate speed and understate friction.
Real adoption
- What it measures
- Share of organisations (and of processes inside each) already using AI in production.
- Examples
- Enterprise surveys (PwC, McKinsey, BCG), platform telemetry (Anthropic, OpenAI, Microsoft Copilot).
- Why it matters
- Separates trial from sustained use. Real adoption is where change starts to be felt.
- If misread
- Counting pilots as real use. Most pilots never reach production without redesigning the process.
Observed labour-market outcomes
- What it measures
- Changes in employment, wages, vacancies, demanded skills and churn attributable to AI.
- Examples
- BLS, ILO, Eurostat, LinkedIn, Indeed, Lightcast, ADP.
- Why it matters
- The only reading that closes the loop between technology and work. Everything else is antecedent.
- If misread
- Attributing to AI structural changes that would have happened anyway.
Evidence vs. hype matrix
Four editorial readings of AI's impact on employment, sorted by evidence strength rather than a numeric score.
Expected labour impact ↑
Evidence strength →
Theoretical exposure
Academic / multilateral methodologies on potential exposure.
IMF — Gen-AI: Artificial Intelligence and the Future of Work
Macroeconomic + labour analysis
IMF — Gen-AI: Artificial Intelligence and the Future of Work
Macroeconomic + labour analysis
ILO — Generative AI and Jobs: A Refined Global Index of Occupational Exposure
Occupational exposure research
ILO — Generative AI and jobs: A 2025 update
Occupational exposure research
Real adoption
Production-use data / actual deployment.
Pending: no verified adoption claim yet.
Observed results
Measured changes in employment, wages or occupations.
Pending: no observed market claim yet.
Weak signals
Surveys, expectations and priority sources without an extracted claim.
World Economic Forum — Future of Jobs Report 2025
Global employer survey
World Economic Forum — Future of Jobs Report 2025
Global employer survey
Stanford HAI — AI Index 2025 Economy chapter
Labour-market data analysis
The matrix implies no causality. A position is not a score.
Verified claims
Verified claims
A small set of source-backed statements with canonical URLs. Each card separates the figure, how to read it, and what the source does NOT prove.
Official source linked: the URL is canonical; the figure is extracted with declared value, date and geography. Concrete figures will be incorporated only when claim, date and methodology are traced.
Tier 1Global employer survey
22 %
of jobs disrupted by 2030
WEF expects labour disruption equivalent to 22% of jobs by 2030, with 170M new roles and 92M displaced, for a net 78M.
- Official source linked
- High confidence
- Global pulse
- IILE component: Labour-market signal
- Geography: Global
- Timeframe: 2030
- How to read it
- Read it as transformation, not destruction: creation, displacement and reshape combined.
- What it does NOT prove
- An employer-survey forecast, not an observed net-employment measurement.
- Source
- World Economic Forum — Future of Jobs Report 2025
Tier 1Global employer survey
39 %
of key skills changed by 2030
Employers expect 39% of the key skills required in the labour market to change by 2030.
- Official source linked
- High confidence
- Executive radar
- IILE component: Skill velocity
- Geography: Global
- Timeframe: 2030
- How to read it
- The minimum-employability bar moves: keeping a job is not enough — the skills mix must update.
- What it does NOT prove
- Intensity varies by sector, country and occupation.
- Source
- World Economic Forum — Future of Jobs Report 2025
Tier 1Macroeconomic + labour analysis
40 %
of global employment exposed
IMF estimates nearly 40% of global employment is exposed to AI.
- Official source linked
- High confidence
- Evidence vs. hype
- IILE component: Technical exposure
- Geography: Global
- Timeframe: 2024
- How to read it
- Exposure measures potential impact, not automatic substitution.
- What it does NOT prove
- Exposure does not equal job loss; some work may be augmented.
- Source
- IMF — Gen-AI: Artificial Intelligence and the Future of Work
Tier 1Macroeconomic + labour analysis
60 %
of advanced-economy jobs exposed
In advanced economies, around 60% of jobs are exposed to AI, mainly because of the prevalence of cognitive work.
- Official source linked
- High confidence
- Evidence vs. hype
- IILE component: Technical exposure
- Geography: Advanced economies
- Timeframe: 2024
- How to read it
- Pressure does not fall only on routine work: cognitive tasks are particularly affected.
- What it does NOT prove
- Complementarity and institutional readiness can change the final outcome substantially.
- Source
- IMF — Gen-AI: Artificial Intelligence and the Future of Work
Tier 1Occupational exposure research
52,558
data points in the methodology
The ILO refined index combines task data, expert input and AI-model predictions; it uses 29,753 tasks, a 1,640-person survey and 52,558 datapoints on automation potential for 2,861 tasks.
- Official source linked
- High confidence
- Methodology
- IILE component: Task transformation
- Geography: Global methodology
- Timeframe: 2025
- How to read it
- The relevant unit for labour impact is the task, not the job alone.
- What it does NOT prove
- The methodology projects exposure/potential; it does not, on its own, measure real adoption or observed outcomes.
- Source
- ILO — Generative AI and Jobs: A Refined Global Index of Occupational Exposure
Tier 1Occupational exposure research
25 %
approx. of jobs exposed to transformation
ILO's 2025 update summarises that roughly one in four occupations may be exposed to transformation by generative AI.
- Official source linked
- Medium confidence
- Global pulse
- IILE component: Task transformation
- Geography: Global
- Timeframe: 2025
- How to read it
- The signal reinforces that the central impact is task transformation, not full automation.
- What it does NOT prove
- Treat it as exposure to transformation, not as direct job loss.
- Source
- ILO — Generative AI and jobs: A 2025 update
Tier 3Labour-market data analysis
The Economy chapter of the AI Index 2025 will be used as a priority source for AI-related labour-demand and skills signals.
- Official source linked
- Medium confidence
- Source radar
- IILE component: Labour-market signal
- Geography: Global
- Timeframe: 2025
- How to read it
- Bridges from theoretical exposure to observable labour-demand signals.
- What it does NOT prove
- Concrete metrics will be incorporated only when extracted and precisely mapped.
- Source
- Stanford HAI — AI Index 2025 Economy chapter
These cards are the first observatory pass with traceable figures. The IILE-IA calibration remains marked editorial / experimental until more verified claims are integrated.
Reading guide
How to read this dashboard
What it is for
To recognise the pressure of labour transformation by sector, occupation and profile — and the capacity to absorb it. A curated editorial reading, not a prediction.
What it is NOT for
It does not predict job destruction. It is not an official statistical index. It does not replace Tier 1–3 sources; it weights them with explicit editorial weights.
How to use it while reading the book
Come back here whenever a chapter cites a specific vector (exposure, adoption, gap) to situate it in the wider picture and compare the rhythm across sectors.
How it will evolve
Future iterations will add real per-source signals, a declared editorial cadence, and an automated watcher for Tier 1–3. IILE-IA calibration will remain editorial.