A study conducted by the Massachusetts Institute of Technology reveals that the threat of automation by artificial intelligence is much greater than previously thought.
By scrutinizing the labor market on an unprecedented scale, researchers estimate that nearly one in eight jobs could already be technically replaceable in the United States. The Yiaho team covers this news.
“The Iceberg Index,” a Novel View of the Labor Market
The core of this discovery relies on a tool called the Iceberg Index, developed in collaboration between MIT and Oak Ridge National Laboratory. This platform digitally recreates the entire American job market by having over 151 million virtual worker profiles interact with artificial intelligence systems.
To achieve this precision, researchers cataloged over 32,000 distinct skills and nearly a thousand different occupations. This ultra-granular approach goes beyond classic studies that focused mainly on technical positions or large cities.
Two Very Different Levels of Reality
The authors clearly distinguish what is visible today from what is technically possible right now. What is called the Surface Index corresponds to current exposure, mainly in the tech sector, and concerns only 2.2% of workers.
The Iceberg Index, on the other hand, reveals the submerged part: 11.7% of American jobs, equivalent to approximately $1.2 trillion in annual payroll, could already be automated with existing tools.
Also read: Can AI Replace Teachers?
The Most Vulnerable Sectors
Contrary to the common belief that only highly skilled IT jobs are threatened, it is primarily administrative positions, business services, finance, and human resources that appear most exposed.
In these fields, a large part of the tasks relies on repetitive processes, structured data analysis, or document processing—operations that artificial intelligence already largely masters.
Source: The Iceberg Index: Measuring Workforce Exposure in the AI Economy
