Methodology
What We Track
This tracker estimates global job losses that are publicly attributed to artificial intelligence, automation, and robotics. We monitor public reporting including news articles, press releases, earnings calls, and company announcements to identify workforce reductions where AI or automation is cited as a contributing factor.
What We Do Not Track
- Job losses with no public reporting or announcement
- Natural attrition or unfilled positions
- Layoffs with no stated connection to AI or automation
- Speculative future predictions (only announced or completed events)
Attribution Model
Each event is classified by the strength of its connection to AI or automation:
Three Estimate Tiers
We present three estimates to reflect uncertainty:
| Category | Conservative | Core | Upper Bound |
|---|---|---|---|
| Explicit | 100% | 100% | 100% |
| Strong | 0% | 75% | 100% |
| Moderate | 0% | 40% | 70% |
| Weak | 0% | 15% | 35% |
| Fringe | 0% | 5% | 15% |
Provisional vs Reviewed Data
Events enter the tracker as provisional when first detected by our automated pipeline. They are later upgraded to reviewed status after deeper analysis confirms or corrects the initial classification. Reviewed records override provisional values. Totals may be revised as better evidence is found.
Deduplication
Multiple news outlets often cover the same event. We group related articles into event clusters and count the underlying event only once, preferring the highest-quality source. Related article links are preserved for reference.
Copyright & Sources
We do not republish full articles. Each event entry includes a short summary generated from the source, along with a link back to the original publisher. Source links remain with original publishers. This site is informational and does not represent official government data.
Corrections
If you believe an event has been misclassified, miscounted, or is missing, please contact us. We actively manage corrections and maintain revision history for transparency.