Artificial intelligence could displace more than one in ten U.S. jobs, according to new research from the Massachusetts Institute of Technology and Oak Ridge National Laboratory. The analysis from the two finds that generative AI systems are already capable of replacing tasks that account for 11.7% of total wages nationwide — a shift that could have fallout across office, retail and multifamily markets if job losses materialize.
The report, developed under MIT’s Project Iceberg, simulates what it calls the “emerging human-AI workforce.” Using MIT’s Large Population Models and Oak Ridge’s Frontier supercomputer, researchers analyzed trillions of data points capturing what 151 million American workers across 923 occupations actually do. They compared those tasks to what today’s AI tools can perform.
Unlike traditional workforce studies, which rely on physical business locations, household surveys or standard occupational categories, this modeling accounts for AI-mediated work — the digital tasks with no geographic anchor or clear job classification. Researchers argue that these hidden activities make traditional indicators unreliable in gauging the accurate scale of AI’s economic impact.
To illustrate the gap, the team created what it calls the Surface Index, which tracks visible AI adoption, such as layoffs in major tech hubs. That measure captures only 2.2% of current job displacement. The Iceberg Index, by contrast, measures what lies below the surface: AI’s ability to take over administrative, financial and professional services tasks nationwide. That “hidden mass” represents the additional 11.7% of jobs and wages exposed to automation.
The model evaluates each occupation’s vulnerability based on three factors — the skills required, how easily those skills can be automated and the economic value of the jobs involved. Based on wage data, visible automation accounts for 2.2% of total income, amounting to about $211 billion annually, while the hidden portion totals roughly $1.2 trillion.
Researchers warn that displacement will likely hit higher-wage, cognitively demanding roles hardest, often without clear replacement opportunities. If those losses occur, the ripple effects could run deep. Curtailed consumer spending could weaken retail real estate, lower household formation might soften multifamily demand, and reduced corporate staffing could lessen office needs.
© Arc, All Rights Reserved. Request academic re-use from www.copyright.com. All other uses, submit a request to TMSalesOperations@arc-network.com. For more information visit Asset & Logo Licensing.