MIT Study: AI Matches 11.7% of US Workforce Tasks

A new study from the Massachusetts Institute of Technology reveals that artificial intelligence systems are already capable of performing tasks equivalent to 11.7 percent of the U.S. workforce, potentially affecting $1.2 trillion in annual wages across sectors like finance, healthcare and professional services. Released this week after research conducted in October, the findings underscore AI's growing technical and economic feasibility in the labor market, though researchers emphasize it does not predict widespread job losses.

The research, part of Project Iceberg developed in collaboration with Oak Ridge National Laboratory, created a digital simulation of the U.S. economy known as the Iceberg Index. This model analyzes 151 million workers, 923 occupations and more than 32,000 skills across 3,000 counties, mapping them against the abilities of over 13,000 existing AI tools. By simulating workers as individual agents with specific locations and expertise, the index assesses where AI can handle cognitive, administrative and routine tasks at costs competitive with or lower than human labor.

Prasanna Balaprakash, a director at Oak Ridge National Laboratory and co-leader of the...