44% of Workers in OECD Countries Face High AI Automation Risk
This isn't about factory workers or truck drivers. The 44% at risk are knowledge workers — accountants, lawyers, analysts, writers — the professional class who thought automation wouldn't reach them.
44%OECD workers in high AI exposure occupations (IMF 2024)
60%Advanced economy jobs with significant AI exposure
~40%Of exposed jobs likely to see wage/hour reduction
~60%Of exposed jobs may see productivity gains (augmentation)
Choose your depth. The data doesn't change — just the explanation.
Almost half of all workers in rich countries have jobs that AI could help do or completely take over. This is different from robots in factories — this is about office jobs. Accountants, lawyers, customer service agents. The people who went to college thinking computers would only replace physical labor are now the ones in the crosshairs.
The IMF's April 2024 World Economic Outlook measured AI "complementarity" and "substitutability" across occupations. For advanced economies: 60% of jobs have significant AI exposure. Of those, roughly 40% may see downward wage/hour pressure as AI substitutes for portions of the role. 60% may see productivity gains without displacement — the "augmentation" outcome. The distribution is not equal: higher-educated, higher-wage workers have MORE exposure but also more ability to adapt. Lower-income emerging market workers have less AI exposure but less adaptability and weaker safety nets.
IMF methodology: O*NET task data × LLM capability profiles × OEWS employment data. "AI exposure" ≠ displacement. The Acemoglu-Restrepo task framework distinguishes automation (AI substitutes human) from augmentation (AI complements human). Key finding: college-educated workers in OECD countries are MORE exposed (60%+) but have higher adaptive capacity. Goldman Sachs estimates 300M full-time equivalent jobs "disrupted" globally. IMF scenario A (optimistic): productivity gains dominate. Scenario B (pessimistic): wage inequality expands significantly, polarization toward high-skill/low-skill with middle-skill hollowing.
Primary: IMF WEO Chapter 3, April 2024 (imf.org/en/Publications/WEO). Methodology appendix: ai_exposure_score = Σ(task_ai_capability × task_weight) × occupation_ai_exposure_weight. Data: O*NET 27.0, OEWS 2023 (bls.gov/oes), BLS Employment Projections. Goldman Sachs: "The Potentially Large Effects of Artificial Intelligence on Economic Growth," Hatzius et al., 2023. Acemoglu framework: "Robots and Jobs: Evidence from US Labor Markets" (NBER WP 23285). Country-level AI exposure tables in IMF WEO Annex 3.A.
Risk by Occupation and Education
The counterintuitive finding: more education = more AI exposure. But higher-educated workers have more adaptability, meaning the risk is split between displacement and augmentation.
AI Automation Risk by Occupation (% of tasks automatable)
IMF World Economic Outlook 2024, O*NET task analysis. Higher = more tasks AI can do today.
AI Exposure by Education Level (OECD)
More education = more AI exposure (counterintuitive).
Displacement vs. Augmentation Split
For exposed workers, two possible outcomes.
The Middle-Class Paradox
Factory automation hit lower-income workers first. AI automation is hitting middle and upper-middle income workers — accountants, paralegals, junior analysts — precisely the jobs that were considered "safe" because they required college degrees. The safety net for this group is weaker than imagined.
Sources
• IMF. "World Economic Outlook: Steady But Slow." Chapter 3: "Gen-AI: Artificial Intelligence and the Future of Work." April 2024.
• Goldman Sachs. "The Potentially Large Effects of Artificial Intelligence on Economic Growth." March 2023.
• BLS. Occupational Employment and Wage Statistics (OEWS) 2023. bls.gov/oes.