The fast rise of artificial intelligence in workplaces has made many workers worry about losing their jobs. A new study from MIT backs up these fears, showing that AI could take over nearly 12% of American jobs—worth about $1.2 trillion in wages.
The Iceberg Index Tool
This research uses the Iceberg Index, a simulation that shows how 151 million workers interact in the job market and how AI might impact them. The index tracks thousands of skills and many professions, helping to see where AI can already replace human workers.
With help from the supercomputer Frontier, the MIT team and Oak Ridge National Laboratory create a ‘virtual lab.’ This lets states and organizations explore ways to retrain workers, reallocate resources, and adapt policies.
Immediate and ‘Hidden’ Exposure
In clear-cut fields like technology, IT, and computing, the risk of job loss is about $211 billion, which is just 2.2% of the total. But most of the risk comes from sectors that often get ignored: accounting, HR, administrative support, financial services, and supply chains.
Overall, researchers believe that the real ‘exposed’ amount reaches $1.2 trillion in wages.
Decentralized Risk
Unlike the common belief that AI will mainly hit big cities, the model shows that risk is spread across all states—even in inland areas not usually talked about in automation discussions.
The Project Iceberg lets each state test economic plans, find high-risk jobs, and create training programs before changes happen in the real economy.
Researchers stress that the aim is not to predict layoffs but to offer a useful planning tool for the age of AI.
