AI Work Index

Theory vs Practice

Where does real-world AI usage diverge most from theoretical exposure? Ranked by the absolute gap between Anthropic's observed AI usage percentile and the theoretical AIOE percentile. Red rows = usage exceeds theory. Blue rows = theory exceeds usage.

# Occupation Gap (pts)DirectionExposureNet Risk Risk Impact
1 Building painter 71311 +75Above theory41%38.3% High At Risk
2 Audiologist 22661 -65Below theory59%11.6% Low Augmented
3 Speech therapist 22662 -65Below theory59%11.7% Low Augmented
4 Library officer 34331 -64Below theory59%12.3% Low Augmented
5 Website administrator/Webmaster 35140 +60Above theory53%27.9% Moderate At Risk
6 Environmental officer (environmental protection) 21331 -59Below theory77%14.2% Low Augmented
7 Landscape architect 21621 -59Below theory55%14.7% Low Augmented
8 IT Infrastructure technician 35121 +59Above theory53%29.3% Moderate At Risk
9 IT security technician 35122 +59Above theory53%28.6% Moderate At Risk
10 IT support technician (including IT user helpdesk technician) 35123 +59Above theory53%28.0% Moderate At Risk
11 Automation engineer (including robotics engineer) 21413 -56Below theory68%10.9% Low Augmented
12 Manufacturing engineer 21411 -56Below theory68%11.3% Low Augmented
13 Process engineer 21415 -56Below theory68%11.5% Low Augmented
14 Production engineer 21412 -56Below theory68%11.2% Low Augmented
15 Quality control/assurance engineer 21414 -56Below theory68%11.5% Low Augmented
16 Chemical engineering technician 31161 +50Above theory54%30.8% High At Risk
17 Chemical engineering technician (petrochemicals) 31163 +50Above theory54%30.3% High At Risk
18 Environmental engineer 21430 -50Below theory78%8.3% Low Augmented
19 Photographer 34310 +48Above theory43%22.9% Moderate Stable
20 Sales supervisor 52201 +48Above theory52%32.2% High At Risk
21 Shop sales assistant 52202 +48Above theory52%32.8% High At Risk
22 Data entry clerk 41320 +47Above theory65%55.2% Very High At Risk
23 Data processing control clerk 43151 +47Above theory65%52.7% Very High At Risk
24 Travel consultant/Reservation executive 42210 +43Above theory38%16.8% Moderate Stable
25 Cabin attendant/steward 51112 +43Above theory38%15.8% Moderate Stable

Gap = Anthropic observed usage percentile minus theoretical AIOE percentile. Positive means more AI adoption than theory predicts. Learn more

Structural AI exposure scores, not employment predictions. Methodology

V3.1 · 2024 wages · Q3 2025 full + Q4 2025 advance labour data · 562 occupations · 80 roles