MIT’s Iceberg Index simulates how today’s AI systems affect 151 million workers across all US states, revealing that only 2.2% of job disruption is visible on the surface—while 11.7% of the workforce and $1.2 trillion in wages are exposed to hidden automation. The model identifies local-level vulnerabilities, helps states test policy scenarios before spending billions, and dispels the myth that AI risks are limited to tech hubs.

AI Is Not Coming for the Workforce — It’s Already Inside It

When MIT publishes a labor study, markets and policymakers tend to sharpen their pencils.
But this time, the numbers landed like a cold plunge.

According to MIT, AI can already replace 11.7% of the US workforce — equivalent to $1.2 trillion in annual wages across finance, professional services, healthcare, HR, logistics, and administrative roles.

Not in theory. Not in 2035.
Already. With existing systems.

That conclusion comes from the Iceberg Index, a new labor-process simulation model created by MIT and Oak Ridge National Laboratory. It analyzes 151 million US workers, 32,000 skills, 923 occupations and 3,000 counties — one of the most granular workforce models ever built.

MIT’s idea is simple:
“If you want to see the future, simulate it before it arrives.”

What Is the Iceberg Index — and Why It Matters Now

The Iceberg Index does not predict when or where job losses will occur. It focuses on something more practical:
What current AI systems can already do relative to human task flows.

The model works as a sandbox allowing policymakers to ask “what if” questions before they spend billions on training programs, workforce subsidies, or AI-related legislation.

It models each worker as an “agent” with:
a set of skills
a location
a profession
a composition of tasks
and a probability that those tasks can be performed by modern AI

Then it simulates how task automation cascades into wage shifts, role redistribution, retraining needs, and regional labor dynamics.

For states like Tennessee, Utah, and North Carolina — already using the tool — this is a GPS for workforce planning rather than a blind guess.

The Hidden Part of the AI Workforce Iceberg ($1.2T)

MIT frames the issue as a literal iceberg.

Above the waterline — the visible part:

layoffs in tech
role consolidation in IT
shifts in engineering workflows

This is only 2.2% of the total wage impact — around $211 billion.

Below the surface — the real shock:
HR routinization
financial operations
logistics coordination
document processing
compliance workflows
administrative chains

These account for nearly $1.2 trillion in exposed wages.

The researchers emphasize that analysts traditionally underestimate these “routine cognitive layers” simply because they are not concentrated in Silicon Valley or New York. Many of the most exposed counties are in Tennessee, Utah, the Carolinas, Texas, Georgia, and the Midwest — regions normally absent in AI discussions.

GEO Insight: AI Exposure Is Not Coastal — It’s National

One of the most disruptive findings is that AI exposure is uniform across all 50 states.

This overturns the long-held belief that:
“AI risk sits mostly in coastal tech hubs.”
According to the Iceberg simulations:
rural counties show the same concentration of at-risk clerical and administrative tasks as major metro areas

Southern states show higher exposure in logistics and HR
Mountain West states show heavy vulnerability in compliance and documentation roles
the Midwest shows outsized exposure in financial back-office operations

Why States Are Rushing to Use Iceberg Before New Laws Pass

Tennessee was the first to officially reference the model in its statewide AI Action Plan.
Utah is preparing to release a report based on it.
North Carolina’s legislators actively use the model for scenario planning.

What’s pulling them in?

Three things:
Local-level detail
Iceberg can zoom into specific counties and ZIP codes, showing exactly which skills are exposed and which training programs would actually matter.

Scenario simulation
Before states spend billions on upskilling, they can test:

What if we subsidize retraining in logistics software?
What if AI tools become 30% more capable next year?
What if we incentivize companies to adopt AI responsibly?

Eliminating blind spots
Traditional economic models rely on surveys and lagging data.
Iceberg shows task-level shifts before they show up in the economy.

MIT researchers note explicitly:
“This is not a forecast model — it is a diagnostic tool showing what AI can already do.”
That distinction matters: the tool isn't predicting collapse; it's identifying cracks before they widen.

A Quiet Revolution in How Labor Policy Will Be Built

AI exposure is no longer a topic for think-tanks and Silicon Valley panels.
States are treating it as infrastructure: something that must be mapped, budgeted, and protected.

Iceberg’s biggest impact may not be in the numbers, but in how it changes policymaking:
targeted retraining instead of generic “learn to code” programs
county-level job transition subsidies
early-warning systems for sudden AI-driven task shifts
smarter allocation of federal workforce funds
stronger alignment between universities and emerging AI workflows

In a sense, the US is building an AI-era FEMA — a preparedness model for economic shocks that haven’t happened yet but are already statistically visible.

The Real Question: Are Companies Prepared as Well as States?

Governments are moving fast.
Businesses — not always.

Finance, healthcare, HR-tech, logistics, and professional services firms face the same $1.2T productivity shock, but many still operate under the assumption that “AI risk is overhyped.”

MIT’s model says otherwise.
If states are planning around AI impacts at ZIP-code precision, companies that ignore task-level automation signals may find themselves hiring, training, and allocating capital with outdated assumptions.

Conclusion
The Iceberg Index paints a simple but uncomfortable picture:
AI isn’t just transforming the economy — it’s rearranging it beneath the surface.

The visible tech layoffs are not the story.
The invisible shift in routine workflows — the trillion-dollar undercurrent — is what will define the next decade of labor markets.
The real advantage now goes to policymakers, firms, and industries that treat simulation not as speculation, but as preparation.

By Miles Harrington 
November 27, 2025

Join us. Our Telegram: @forexturnkey
All to the point, no ads. A channel that doesn't tire you out, but pumps you up.