According to Fortune, MIT researchers have determined that current AI systems can already perform work equivalent to 11.7% of the U.S. labor market. The study, written in October but released this week, estimates this represents about 151 million workers and roughly $1.2 trillion in annual wages. Unlike previous research that focused on theoretical automation exposure, this analysis specifically looks at jobs where AI can perform tasks at a cost competitive with or cheaper than human labor. The findings come from Project Iceberg, a large-scale labor simulation developed by MIT with Oak Ridge National Laboratory. The model creates what researchers call a “digital twin of the U.S. labor market,” tracking over 32,000 skills across 923 job types in 3,000 counties.
How the Iceberg model works
Here’s what makes this research different from all those “AI will take your job” headlines we’ve seen before. The MIT team didn’t just look at whether AI could do certain tasks – they built an incredibly detailed simulation that models whether it makes economic sense to actually replace humans. We’re talking about simulating 151 million individual workers as agents, each with specific skills, occupations, and locations. Basically, they created a massive virtual lab where they could test how AI would perform across the entire U.S. economy. The model maps what current AI systems can actually do against real-world job requirements and costs. And that’s why these numbers carry more weight than previous estimates.
The white-collar reality check
Now here’s the really interesting part. While most of the public AI discussion has been about tech jobs and coding, the study reveals something much broader. Current AI adoption is concentrated in tech work, representing about $211 billion in wages. But the researchers found AI is already capable of handling cognitive and administrative tasks across finance, healthcare, and professional services that represent around $1.2 trillion in wages. That’s five times the currently visible impact. So basically, we’re only seeing the tip of the iceberg when it comes to AI’s potential disruption. Finance, healthcare administration, HR, logistics, legal, and accounting work – these are the areas where existing AI tools can already execute many routine tasks. Pretty sobering for anyone who thought their white-collar job was safe.
But will jobs actually disappear?
Here’s where we need to be careful about jumping to conclusions. The MIT researchers are very clear that technical capability doesn’t automatically mean widespread job losses. Earlier research from MIT’s Computer Science and Artificial Intelligence Laboratory found that fully replacing human workers with AI remains too expensive or impractical in many cases, even where the technology could perform the tasks. And separate research from MIT Sloan concluded that from 2010 to 2023, AI exposure didn’t lead to broad net job losses and often coincided with faster revenue and employment growth at adopting firms. So what’s really happening? Companies are using AI to augment workers rather than replace them entirely. The technology is changing job descriptions rather than eliminating positions wholesale.
business-and-government”>What this means for business and government
The Iceberg Index isn’t designed to predict specific layoffs. Instead, it gives policymakers and business leaders a way to stress-test different scenarios before committing training dollars or infrastructure spending. States like Tennessee, North Carolina and Utah are already using the platform to evaluate how AI might reshape their workforces. For companies, the message is clear: the window to treat AI as a distant future issue is closing fast. And for governments, this raises urgent questions about retraining workers, supporting high-exposure regions, and adapting social safety nets. The full MIT report and Project Iceberg platform provide the detailed analysis behind these findings. The bottom line? We’re not looking at immediate mass unemployment, but we are looking at the most significant workforce transformation in generations – and it’s already happening.
