In 2025, artificial intelligence has advanced to a level where it can already technically perform the work equivalent to 11.7% of the U.S. labor market, representing approximately $1.2 trillion in wages. This finding emerges from the Iceberg Index, a labor simulation tool developed by the Massachusetts Institute of Technology and Oak Ridge National Laboratory, designed to provide detailed insight into AI’s overlap with human job skills across the country. Unlike conventional studies that focus on visible job losses after automation happens, the Iceberg Index offers a forward-looking view on where AI capabilities lie beneath the surface, mapping potential workforce exposure down to zip code level.
The Iceberg Index models 151 million U.S. workers as individual agents, considering thousands of unique skills across hundreds of occupations and thousands of counties. It assesses what parts of these occupations AI systems can already perform, identifying not just displaced jobs but the underlying technical exposure. This perspective is crucial because AI disruption is far broader than what is immediately visible in high-profile sectors or tech-heavy cities. In fact, the study highlights that AI’s reach is geographically diverse, touching urban, suburban, and rural areas alike, contrary to the common assumption that AI impacts only thrive in coastal hubs.
For example, states in the Rust Belt such as Ohio, Michigan, and Tennessee reveal modest AI exposure when only current technology adoption is considered, but dramatically higher exposure emerges when cognitive and administrative tasks, areas ripe for AI automation, are factored in. These tasks cut across manufacturing-related positions and service roles often overlooked in AI conversations. The index underscores that such indirect and supportive roles form a substantial portion of AI’s technical footprint, reflecting a ripple effect that goes beyond frontline automation.
One of the Iceberg Index’s most powerful contributions is its ability to offer actionable intelligence for policymakers. By providing a granular map of workforce exposure by location and skill type, local and state governments can make informed decisions about where to invest billions in reskilling and infrastructure. Tennessee and Utah, among others, have already begun using the Iceberg Index as a foundation for their AI workforce strategies, developing task forces and working groups that leverage this data to design targeted training and economic development policies. This tool enables them to simulate scenarios before committing resources, thereby mitigating risks and capitalizing on opportunities with greater precision.
The study also stresses an important caution: technical exposure does not necessarily translate directly to layoffs or job losses. AI’s actual impact depends on how businesses, workers, and governments respond in real time. Adaptation, reorganization, and complementary human-AI collaboration are critical variables that will determine how the labor market evolves. The Iceberg Index thus serves as a strategic compass to prepare for a manageable transition rather than a prediction of inevitable job destruction.
Looking forward, the Iceberg Index is pivotal for reshaping workforce development initiatives. It shifts the narrative from reactive to proactive by highlighting AI’s underappreciated diffusion into a wide range of occupations and communities. This insight calls for policies that not only focus on training for high-tech skills but also prioritize cognitive and administrative capabilities that support multiple industries. Ultimately, this new model of workforce exposure could inform smarter investments in education, economic diversification, and infrastructure that align with the complex realities of AI’s integration into work across the nation.
