MIT Study Reveals AI Replaces 11.7 Percent of US Workforce
Artificial intelligence systems now perform tasks equivalent to 11.7 percent of the US labor market, exposing $1.2 trillion in annual wages to automation. This reaches beyond tech sectors into finance, health care, and professional services, where routine functions face immediate displacement risks. The analysis challenges assumptions that AI impacts remain confined to coastal hubs, showing vulnerabilities across all 50 states and rural counties.
Researchers at the Massachusetts Institute of Technology and Oak Ridge National Laboratory developed the Iceberg Index, a simulation modeling 151 million US workers as individual agents tagged with skills, tasks, occupations, and locations. The tool maps over 32,000 skills across 923 occupations in 3,000 counties, assessing where current AI capabilities match human functions. It simulates AI-human workforce interactions, revealing labor flows and task reshuffling before economic shifts appear in data.
Visible AI adoption in computing and information technology accounts for just 2.2 percent of wage exposure, or $211 billion, mainly through layoffs and role changes. Hidden risks dominate in human resources, logistics, finance, and office administration, where AI handles repetitive processes like data entry and scheduling. Exposed occupations span urban and rural areas, with simulations projecting broader productivity gains if reskilling aligns with these patterns.
Finance emerges as highly susceptible, with AI automating compliance checks, risk assessments, and transaction processing across banks and investment firms. Health care faces disruption in administrative billing and patient record management, though physical care roles provide buffers. Professional services, including legal research and consulting analytics, see AI agents drafting reports and summarizing documents at scale.
The Iceberg Index enables state-level policy testing through interactive scenarios, adjusting variables like training budgets and technology rollout speeds to forecast employment and GDP effects. Tennessee incorporated the model into its November 2025 AI Workforce Action Plan, prioritizing investments in health care and manufacturing. North Carolina and Utah followed with validations using local labor data, simulating interventions down to census blocks.
Prasanna Balaprakash, Oak Ridge National Laboratory director and project co-leader, described the Index as “a digital twin for the US labor market” that uncovers pre-market shifts in tasks and skills. North Carolina state Senator DeAndrea Salvador highlighted its granularity, noting it reveals “county-specific data on skills automation and impacts on state GDP and employment.” These tools position states to target reskilling for 12 million workers in high-exposure roles.
Implementation focuses on augmentation over replacement, integrating AI with robotics in physical sectors like transportation and nuclear energy. Tennessee officials emphasized using AI assistants to bolster rather than erode industries reliant on hands-on work. The model runs population-scale experiments, projecting that policy tweaks could mitigate up to 30 percent of displacement effects in vulnerable regions.
Overall, the study equips policymakers with granular forecasts, emphasizing proactive measures like $1 billion reskilling funds to redirect exposed workers into AI-complementary roles. By simulating adoption curves, it anticipates labor reallocations that traditional metrics overlook, informing federal guidelines on workforce transitions. This framework underscores AI’s role in amplifying human productivity while demanding structured safeguards against uneven economic fallout.
