A new study from the Massachusetts Institute of Technology (MIT) shows that AI may be poised to replace far more jobs than previously predicted. A large amount of hidden data reveals that artificial intelligence could now occupy 11.7% of the labor market, according to researchers.
The new estimates come courtesy of a project called the Iceberg Index, created through a partnership between MIT and Oak Ridge National Laboratory (ORNL), a federally funded research center in Tennessee. According to its website, the Iceberg Index “simulates an agent-based United States, a human and AI workforce of more than 151 million human workers working alongside thousands of AI agents.” Simply put, this tool is designed to accurately simulate how AI is poised to disrupt the current workforce, down to the zip code of a specific region.
The Iceberg Index model treats America’s more than 151 million workers as individual agents, each categorized by skill, task, occupation, and location. In total, more than 32,000 skills and 923 occupations have been mapped across 3,000 counties. In an interview with CNBC, ORNL Director and study co-lead Prasanna Varaprakash described it as a “digital twin of the U.S. labor market.” The index uses its database to analyze how well digital AI tools can already perform certain technical and cognitive tasks and to estimate what the AI exposure in each area will be.
Already, state governments in Tennessee, North Carolina, and Utah are using the index to prepare for AI-driven workforce changes. Here are three key takeaways from the study.
AI is more pervasive in the workforce than we think
Perhaps the study’s biggest finding is the discovery of what it calls a “substantive measurement gap” in how we typically think about AI replacing jobs.
According to the report, if analysts looked only at current AI deployments (mainly concentrated in computing and technology), they would find that exposure to AI amounts to only about 2.2% of the workforce, or about $211 billion in wages. (The report calls this the “surface index.”) But it also says it’s “just the tip of the iceberg.”
When you factor in variables such as the potential for AI automation in government, finance, and professional services, that number rises to 11.7% of the workforce and approximately $1.2 trillion in wages. (This calculation is called the “iceberg index.”)
The study’s authors stress that these results only represent technical AI exposure and not actual future evacuation outcomes. These will depend on how businesses, workers and local governments adapt over time.
AI takeover isn’t limited to beaches
It is fairly common to think that the most jobs exposed to AI will be concentrated in coastal hubs, where technology companies are primarily concentrated. However, the Iceberg Index shows that AI’s ability to take over tasks from employees is much more widely distributed.
According to the study, many U.S. states record low AI impacts when considering only current AI adoption in computing and technology, but much higher values when other variables are taken into account.
“Rust Belt states such as Ohio, Michigan, and Tennessee record modest surface index values but significant iceberg index values driven by cognitive tasks such as financial analysis, management coordination, and professional services that support manufacturing operations,” the study states.
What difference does this data actually make?
Now that MIT and ORNL have successfully established the Iceberg Index, they hope local governments can use it to protect workers and the economy. Local councilors can use this map to gain granular insights, such as looking at specific city blocks to see which skill sets are most used and the potential for automation.
According to CNBC, MIT and ORNL also built an interactive tool that allows states to try different policy measures, such as adjusting training programs or changing workforce spending, and predict how those changes will affect local jobs and gross domestic product.
“The Iceberg Index provides measurable intelligence for key workforce decisions, including where to invest in training, which skills to prioritize, and how to balance infrastructure and human capital,” the report states. “This index reveals not only the visible disruptions in the technology sector, but also the larger transformations that are taking place behind the scenes. By measuring exposure before adoption reshapes work, the index enables countries to prepare rather than react, turning AI into a navigable transition.”
The final deadline for Fast Company’s World Changing Ideas Awards is Friday, December 12th at 11:59pm PT. Apply now.
