Particle.news
Download on the App Store

MIT’s ‘Project Iceberg’ Finds AI Can Do 11.7% of U.S. Work, Outpacing Adoption Fivefold

A labor-market digital twin built with Oak Ridge data models 151 million workers to guide targeted planning.

Overview

  • The model puts technically feasible AI work at 11.7% of labor, or about $1.2 trillion in wages, describing capability rather than an immediate layoff forecast.
  • Real-world use remains concentrated in software work worth roughly 2.2% of wages, about $211 billion, creating a fivefold gap with capability.
  • Exposure clusters in routine cognitive tasks in finance, healthcare administration, human resources, logistics, and professional services such as legal and accounting.
  • The study finds vulnerability dispersed across all states, not just coastal tech hubs, signaling that regions with small tech sectors may be overlooking risk.
  • States including Tennessee, North Carolina, and Utah are adopting the Iceberg Index to test scenarios, identify hotspots, and inform retraining and investment plans.