The AI Anxiety Index
Media hype vs. actual job displacement risk. We cross-referenced 1,472 tagged article-occupation mentions from 124 news sources with our task-level AI risk scores and BLS employment data.
Key finding: The media fixates on software developers (76 mentions, 35% risk), but occupations like office clerks and medical records technicians — representing millions more workers — face higher actual displacement risk with a fraction of the coverage.
How to read this chart: Each bubble is an occupation. Horizontal position = AI displacement risk (higher = more at risk). Vertical position = media attention (higher = more coverage). Bubble size = number of workers employed (BLS data). Color = salary band.
The dashed lines show the median — jobs in the top-left are overhyped (lots of coverage, lower risk), while jobs in the bottom-right are underreported (high risk, little coverage).
Most Overhyped
Jobs getting disproportionate media attention relative to their actual risk score.
Most Underreported
Jobs with high actual risk but relatively little media coverage.
Methodology
Media mentions are counted from articles collected by our news pipeline across 124 sources (Bloomberg, TechCrunch, WSJ, healthcare and legal trade press, etc.), filtered for relevance to AI and workforce impact, then tagged to O*NET occupations using hybrid keyword + Gemini classification.
Risk scores are computed from individual task-level automation assessments for each occupation, discounted by protective factors (social intelligence, regulatory barriers, physical dexterity, etc.). See our full methodology.
Employment and salary data from the U.S. Bureau of Labor Statistics Occupational Employment and Wage Statistics (OEWS) program.
Only occupations with 3+ article mentions are shown. Bubble size uses BLS employment counts where available; occupations without BLS data use a default size.