Which Careers Are Most Exposed to AI?
Compare 873+ occupations by AI risk, salary, and typical education to find future-proof career paths.
Methodology: How We Calculate AI Risk
We used the Pew Research Center's AI exposure methodology to categorize occupations by AI risk.
1. Data Foundation
- • Source: O*NET Work Activities database
- • Coverage: 873 U.S. occupations
- • Ratings: Pew's activity scoring rubric. Each activity scored on a 1–5 Importance scale (1 = "not important," 5 = "extremely important")
2. How Pew Classify's AI Exposure
- • High Exposure (16 activities): Tasks AI can already perform (e.g. information processing, data analysis, documentation)
- • Medium Exposure (16): Tasks where AI offers assistance (e.g. customer support, training/teaching, coordination)
- • Low Exposure (9): Tasks demanding human presence or fine motor skills (e.g. manual repair, equipment operation, physical activities)
3. How We Categorize Occupations and High, Medium or Low Risk
1. Compute Raw Averages:
- • For each occupation, take the average Importance (1–5) of its 16 High-exposure activities → Avg_High
- • Do the same for its 9 Low-exposure activities → Avg_Low
2. Determine Cutoffs:
- • Find the 75th percentile of all occupations' Avg_High scores → High-risk cutoff
- • Find the 75th percentile of all occupations' Avg_Low scores → Low-risk cutoff
3. Label Risk:
- • High Risk: occupations with Avg_High ≥ High-risk cutoff
- • Low Risk: occupations with Avg_Low ≥ Low-risk cutoff
- • Medium Risk: all others