Artificial intelligence systems are now powerful and inexpensive enough to perform work equivalent to nearly 12 percent of U.S. jobs, according to a new MIT study — a finding likely to intensify pressure on employers, workers and policymakers to prepare for rapid changes in the economy.
The research, written in October and released Wednesday, estimates that current AI tools could already take over tasks tied to 11.7 percent of the U.S. labor market. That represents roughly 151 million workers and about $1.2 trillion in annual wages. Unlike earlier research that looked broadly at “exposure” to automation, the MIT analysis focuses on tasks that AI can perform at a cost that is competitive with, or cheaper than, human labor.
The findings come from Project Iceberg, a large-scale labor-market simulation developed by MIT in collaboration with Oak Ridge National Laboratory, home to the Frontier supercomputer. The model creates what researchers call a “digital twin of the U.S. labor market,” simulating millions of workers as individual agents with specific skills, occupations and locations. It maps more than 32,000 skills across 923 job types in 3,000 counties and compares them with existing AI capabilities.
“We’re effectively creating a digital twin of the U.S. labor market,” Prasanna Balaprakash, a director at Oak Ridge National Laboratory and co-leader of the study, told CNBC.
Capability is not the same as job loss
The study stresses that the 11.7 percent figure reflects technical capability and economic feasibility, not a forecast of immediate job displacement. For now, AI adoption remains concentrated in tech roles — primarily coding — representing about 2.2 percent of wage value, or $211 billion. But the analysis shows that AI is already capable of performing cognitive and administrative tasks across finance, healthcare, human resources, logistics and professional services that amount to $1.2 trillion in wages.
That means much of the potential disruption sits in white-collar, back-office work rather than the manual or frontline jobs that have dominated public conversations.
Even so, MIT researchers caution that capability does not guarantee widespread replacement. Previous research from MIT’s Computer Science and Artificial Intelligence Laboratory found that, in many sectors, fully substituting AI for human labor remains too costly or impractical. A separate MIT Sloan analysis of 2010 to 2023 found that AI exposure often coincided with increased revenue and employment growth for firms adopting the technology.
A planning tool for states and employers
The Iceberg Index is not meant to predict layoffs. Instead, it gives policymakers and employers a way to model different scenarios before they make decisions about retraining programs, regulatory changes or investment in infrastructure. Tennessee, North Carolina and Utah have already begun using the platform to assess how AI may reshape their workforces.
For companies, the researchers say, the findings show that the window to treat AI as a long-term issue is closing. For governments, the study raises questions about how to retrain workers, support highly exposed industries and update tax and social-protection systems for a labor market where software can already perform a meaningful share of work.
