Source policy
Every public claim should name the source, geography, publication date, data vintage, and limitation. Model scores should carry a version and confidence label so readers can distinguish observation from inference.
- World Economic Forum: Survey-based projections reflect employer expectations, not realized displacement.
- International Labour Organization: Exposure estimates do not directly predict layoffs or wage outcomes.
- McKinsey Global Institute: Scenario analysis depends on adoption timing, demand growth, and occupational transition assumptions.
- OECD: Automation-risk framing varies by country, institution, and workplace implementation.
- US Bureau of Labor Statistics: Wage and employment figures are historical labor-market measures, not AI-specific forecasts.
- O*NET: Task descriptions require modeling to translate into AI exposure and displacement pressure.