Springfield Parks & Rec

Measuring U.S. Workers' Capacity to Adapt to AI-Driven Job Displacement

A comprehensive analysis of occupational exposure, adaptive capacity, and geographic vulnerability across the American workforce

By Sarah Chen, Michael Torres, Priya Anand

Introduction

Artificial intelligence is rapidly transforming the American labor market. While AI promises significant productivity gains, it also raises pressing questions about workforce displacement. Not all workers face equal risk — and not all are equally prepared to adapt.

This analysis measures two dimensions for 25 major occupations: AI exposure (how much of the job's core tasks can be augmented or replaced by current AI systems) and adaptive capacity (the worker's ability to transition to new roles based on education, skill transferability, and labor market demand).

Our findings reveal a workforce divided into four distinct quadrants — and a geography of vulnerability that maps onto America's economic fault lines.

Approach and Methodology

We scored each occupation on two axes. AI exposure (0 to 1) measures the share of core tasks susceptible to automation based on current large language model and machine learning capabilities. Adaptive capacity (0 to 1) captures transferable skills, educational attainment, cross-industry demand, and historical labor mobility patterns.

Employment figures are drawn from the Bureau of Labor Statistics Occupational Employment and Wage Statistics for 2025. Metro-level vulnerability concentrations are modeled by mapping occupational employment shares to local economies.

AI exposure vs. adaptive capacity across occupations

Bubble size represents total employment (millions)

Bubble size represents total employment (millions) Scatter plot.0.150.010.320.20.480.390.650.570.810.760.980.95Software Developers: ai_exposure=0.82, adaptive_capacity=0.91Registered Nurses: ai_exposure=0.34, adaptive_capacity=0.78Office Clerks: ai_exposure=0.76, adaptive_capacity=0.42Accountants: ai_exposure=0.85, adaptive_capacity=0.67Truck Drivers: ai_exposure=0.12, adaptive_capacity=0.31Teachers K-12: ai_exposure=0.41, adaptive_capacity=0.73Lawyers: ai_exposure=0.73, adaptive_capacity=0.88Retail Salespersons: ai_exposure=0.28, adaptive_capacity=0.35Physicians: ai_exposure=0.45, adaptive_capacity=0.94Data Entry Keyers: ai_exposure=0.91, adaptive_capacity=0.29Financial Analysts: ai_exposure=0.87, adaptive_capacity=0.82Customer Service Reps: ai_exposure=0.72, adaptive_capacity=0.38Mechanical Engineers: ai_exposure=0.56, adaptive_capacity=0.86Janitors: ai_exposure=0.05, adaptive_capacity=0.22Marketing Managers: ai_exposure=0.68, adaptive_capacity=0.84Paralegals: ai_exposure=0.82, adaptive_capacity=0.54Graphic Designers: ai_exposure=0.74, adaptive_capacity=0.63Construction Laborers: ai_exposure=0.08, adaptive_capacity=0.28Pharmacists: ai_exposure=0.61, adaptive_capacity=0.79Executive Assistants: ai_exposure=0.79, adaptive_capacity=0.47Web Developers: ai_exposure=0.76, adaptive_capacity=0.89Security Guards: ai_exposure=0.15, adaptive_capacity=0.26Management Analysts: ai_exposure=0.71, adaptive_capacity=0.87Food Service Workers: ai_exposure=0.11, adaptive_capacity=0.19HR Specialists: ai_exposure=0.65, adaptive_capacity=0.72
Download CSV

Key Findings

The four quadrants

The scatter plot above reveals four distinct groups:

High exposure, high capacity (upper right) — Occupations like software developers, lawyers, and financial analysts. These workers face significant AI disruption but possess strong adaptive skills. They are most likely to evolve alongside AI rather than be replaced by it.

High exposure, low capacity (lower right) — Office clerks, data entry keyers, and customer service representatives. This is the most vulnerable group: their tasks are highly automatable, and their skill profiles offer fewer transition pathways.

Low exposure, high capacity (upper left) — Registered nurses, teachers, and physicians. These roles require interpersonal skills and physical presence that limit AI substitution.

Low exposure, low capacity (lower left) — Janitors, food service workers, and construction laborers. While safe from AI displacement, these workers face other labor market challenges including wage stagnation.

Occupation rankings by adaptive capacity

Occupation rankings by adaptive capacity
occupation adaptive_score employment median_wage ai_exposure_pct
Physicians 94.2 730000 229300 45
Software Developers 91.3 1870000 127260 82
Lawyers 88.1 810000 145760 73
Mechanical Engineers 86.4 310000 96310 56
Management Analysts 87.0 980000 99410 71
Marketing Managers 84.2 310000 140040 68
Web Developers 89.0 210000 80730 76
Financial Analysts 82.5 330000 95080 87
Pharmacists 79.3 330000 132750 61
HR Specialists 72.1 780000 67650 65
Registered Nurses 78.0 3180000 81220 34
Teachers K-12 73.4 3660000 61690 41
Accountants 67.2 1440000 79880 85
Graphic Designers 63.1 270000 57990 74
Paralegals 54.3 350000 59200 82
Executive Assistants 47.1 580000 65980 79
Office Clerks 42.0 2540000 38030 76
Customer Service Reps 38.4 2970000 37780 72
Retail Salespersons 35.2 3690000 31920 28
Truck Drivers 31.0 2010000 49920 12
Data Entry Keyers 29.1 160000 36970 91
Construction Laborers 28.3 1290000 40750 8
Security Guards 26.0 1160000 35300 15
Food Service Workers 19.4 3450000 27620 11
Janitors 22.1 2370000 31990 5
Download CSV

Geographic Distribution of Vulnerability

AI vulnerability is not evenly distributed across the country. Metro areas with heavy concentrations of clerical, administrative, and routine-cognitive occupations face the greatest displacement risk. The national average vulnerability rate is 3.9% of the local workforce.

College towns and tech hubs (San Jose at 2.4%, Austin at 2.8%) show the lowest vulnerability, reflecting their concentration of high-adaptive-capacity workers. Meanwhile, metros in the Sun Belt and parts of the Midwest — San Antonio (5.8%), Houston (5.2%), Jacksonville (5.1%) — face disproportionate risk due to large service-sector and administrative workforces.

Metro-area AI vulnerability concentration

Percentage of local workforce in high-risk, low-adaptive-capacity occupations

Interactive US metro map showing vulnerability_pct by city. Alabama Arizona Arkansas California Colorado Connecticut Delaware District of Columbia Florida Georgia Idaho Illinois Indiana Iowa Kansas Kentucky Louisiana Maine Maryland Massachusetts Michigan Minnesota Mississippi Missouri Montana Nebraska Nevada New Hampshire New Jersey New Mexico New York North Carolina North Dakota Ohio Oklahoma Oregon Pennsylvania Rhode Island South Carolina South Dakota Tennessee Texas Utah Vermont Virginia Washington West Virginia Wisconsin Wyoming New York, NY: 2% Los Angeles, CA: 6.2% Chicago, IL: 7% Houston, TX: 2.7% Phoenix, AZ: 2.2% Philadelphia, PA: 4.8% San Antonio, TX: 5.3% San Diego, CA: 3.3% Dallas, TX: 6.3% San Jose, CA: 2.4% Austin, TX: 3.4% Jacksonville, FL: 2.3% Fort Worth, TX: 3% Columbus, OH: 2.9% Indianapolis, IN: 2.7% Charlotte, NC: 2.5% San Francisco, CA: 2.4% Seattle, WA: 3.1% Denver, CO: 3% Washington, DC: 3.1% Nashville, TN: 3.4% Oklahoma City, OK: 3.7% El Paso, TX: 3.2% Boston, MA: 4.3% Portland, OR: 2.6% Las Vegas, NV: 4.5% Memphis, TN: 4.7% Louisville, KY: 5.7% Baltimore, MD: 4.2% Milwaukee, WI: 5.9% Albuquerque, NM: 2.5% Tucson, AZ: 2.5% Fresno, CA: 3% Sacramento, CA: 6% Kansas City, MO: 5.2% Mesa, AZ: 2.8% Atlanta, GA: 3.6% Omaha, NE: 4.6% Colorado Springs, CO: 4.9% Raleigh, NC: 2.4% Long Beach, CA: 4.5% Virginia Beach, VA: 5.2% Miami, FL: 5.3% Oakland, CA: 3% Minneapolis, MN: 2.6% Tulsa, OK: 4.2% Bakersfield, CA: 2.7% Wichita, KS: 6.2% Arlington, TX: 2.6% Aurora, CO: 5% Tampa, FL: 3% New Orleans, LA: 6.4% Cleveland, OH: 3.8% Honolulu, HI: 6% Anaheim, CA: 5.1% Lexington, KY: 3.3% Stockton, CA: 6.6% Henderson, NV: 6.4% Corpus Christi, TX: 3.7% Riverside, CA: 6.7% Newark, NJ: 6.5% St. Paul, MN: 6.4% Santa Ana, CA: 3.6% Cincinnati, OH: 4.9% Irvine, CA: 3.3% Orlando, FL: 2.7% Pittsburgh, PA: 4.1% Greensboro, NC: 6.4% Jersey City, NJ: 3.8% Anchorage, AK: 5.4% Lincoln, NE: 3.1% Plano, TX: 3.2% Durham, NC: 3.4% Buffalo, NY: 2.8% Chandler, AZ: 2.3% Chula Vista, CA: 7% Madison, WI: 3.3% Gilbert, AZ: 3% Reno, NV: 2.3% Lubbock, TX: 3.5% North Las Vegas, NV: 2.6% St. Petersburg, FL: 3.7% Irving, TX: 5.1% Laredo, TX: 4.1% Winston-Salem, NC: 6.9% Chesapeake, VA: 3.3% Glendale, AZ: 2.1% Garland, TX: 4.5% Scottsdale, AZ: 3.1% Norfolk, VA: 2.7% Boise, ID: 2.4% Fremont, CA: 2.9% Spokane, WA: 2.4% Santa Clarita, CA: 6.6% Baton Rouge, LA: 4.3% Richmond, VA: 5.9% Hialeah, FL: 2.1% San Bernardino, CA: 5.2% Tacoma, WA: 6.6% Modesto, CA: 4.7% Huntsville, AL: 2.7% Fontana, CA: 5.7% Des Moines, IA: 5.7% Moreno Valley, CA: 3% Fayetteville, NC: 4.4% Yonkers, NY: 4% Rochester, NY: 5.3% Columbus, GA: 3% Worcester, MA: 5.8% Port St. Lucie, FL: 3.2% Little Rock, AR: 4.7% Augusta, GA: 2.2% Oxnard, CA: 2.1% Birmingham, AL: 6.3% Montgomery, AL: 6.4% Frisco, TX: 2.7% Amarillo, TX: 4.6% Salt Lake City, UT: 2.6% Grand Rapids, MI: 2.6% Overland Park, KS: 4.4% Tallahassee, FL: 3.1% Grand Prairie, TX: 6.7% McKinney, TX: 6.1% Cape Coral, FL: 5.6% Sioux Falls, SD: 3.2% Peoria, AZ: 4% Providence, RI: 4.2% Vancouver, WA: 2.5% Knoxville, TN: 3.8% Akron, OH: 2.3% Brownsville, TX: 5.3% Newport News, VA: 5.1% Fort Lauderdale, FL: 2.3% Chattanooga, TN: 5% Tempe, AZ: 3.1% Rancho Cucamonga, CA: 5.5% Eugene, OR: 6.2% Oceanside, CA: 5.7% Elk Grove, CA: 5% Salem, OR: 4.4% Ontario, CA: 4.6% Cary, NC: 3.3% Lancaster, CA: 6.4% Garden Grove, CA: 6.8% Pembroke Pines, FL: 2.1% Fort Collins, CO: 5.1% Palmdale, CA: 3.6% Springfield, MO: 5.1% Clarksville, TN: 5.3% Hayward, CA: 4.6% Paterson, NJ: 3.8% Alexandria, VA: 4.4% Macon-Bibb County, GA: 5.1% Corona, CA: 3.2% Lakewood, CO: 2.6% Springfield, MA: 5.1% Sunnyvale, CA: 3% Jackson, MS: 2.2% Killeen, TX: 2.7% Murfreesboro, TN: 6.6% Pasadena, TX: 3.1% Bellevue, WA: 2.7% Joliet, IL: 5.9% Charleston, SC: 4.2% Mesquite, TX: 6.8% Naperville, IL: 6.1% Rockford, IL: 5.1% Syracuse, NY: 3.9% Bridgeport, CT: 6.6% Savannah, GA: 5.2% Roseville, CA: 3.6% Midland, TX: 2.3% Torrance, CA: 2% Surprise, AZ: 5.8% McAllen, TX: 4% Visalia, CA: 4.5% Olathe, KS: 5.8% Gainesville, FL: 6.6% West Valley City, UT: 3.4% Denton, TX: 3.5% Waco, TX: 6.3% Pasadena, CA: 3.3% Cedar Rapids, IA: 3.1% Dayton, OH: 4.7% Elizabeth, NJ: 2.9% Hampton, VA: 6.7% Columbia, SC: 4.4% Stamford, CT: 2.2% Miramar, FL: 4.2% Coral Springs, FL: 5.4% Sterling Heights, MI: 2.5% New Haven, CT: 4.2% Concord, CA: 5.2% Topeka, KS: 5.7% Thousand Oaks, CA: 2.2% Simi Valley, CA: 6.5% Fargo, ND: 6.5% Arvada, CO: 2.5% Wilmington, NC: 7% Hartford, CT: 4.3% College Station, TX: 4.6% Palm Bay, FL: 4.1% Round Rock, TX: 2.3% Meridian, ID: 5.8% Clearwater, FL: 2.3% West Palm Beach, FL: 3.4% Evansville, IN: 3.3% Westminster, CO: 5% League City, TX: 5.3% Provo, UT: 2.3% Lakeland, FL: 3% Pompano Beach, FL: 5% Pueblo, CO: 6.6% Lewisville, TX: 6.4% Centennial, CO: 2.5% Santa Maria, CA: 3.6% Sparks, NV: 4.7% Tyler, TX: 5.4% Green Bay, WI: 4.6% Davie, FL: 6.5% Tuscaloosa, AL: 3.6% Nampa, ID: 3.5% Bend, OR: 5.2% Goodyear, AZ: 4.5% San Marcos, TX: 4.7%
vulnerability_pct (%)
2
7
Download CSV

Implications for Policy

Our analysis suggests three priorities for policymakers and employers:

Targeted retraining programs. The 8.5 million workers in high-exposure, low-capacity occupations need accessible upskilling pathways — particularly in digital literacy, data analysis, and interpersonal service roles that remain AI-resistant.

Place-based investment. Geographic concentration of vulnerability means that some metro areas will bear a disproportionate share of labor market disruption. Federal and state programs should direct transition support to the most affected regions.

Adaptive infrastructure. Employers should invest in continuous learning systems rather than one-time retraining. The pace of AI advancement means today's safe occupations may face new exposure within five to ten years.

Conclusion

The AI transformation of work is not a distant prospect — it is underway. Our analysis shows that while a majority of American workers have the adaptive capacity to navigate this transition, a significant minority face compounding disadvantages: high task exposure, limited transferable skills, and concentration in metro areas already under economic strain.

Understanding where workers fall on the exposure-capacity spectrum is the first step toward ensuring that AI-driven productivity gains are broadly shared rather than concentrated among those already best positioned to benefit.