Browse exposure scores from four sources: OpenAI AI-rated & human-rated tiers, Brynjolfsson et al. SML scores, and Anthropic behavioral data (task success, AI autonomy, work use). Switch sources using the dropdown.
Each bar is a sector's average exposure score. Sectors at the top have the most AI-performable tasks today. Click any bar to filter the whole page to that sector.
Counts how many occupations fall in each exposure band. The shaded overlay is your current filter, so you can see where a sector or search result sits versus all 923 jobs.
Tests the "will AI replace high-paid work?" question. The cloud tilts slightly upward — higher-exposure jobs tend to pay more, because exposure tracks cognitive, knowledge-based tasks rather than low-wage manual ones.
Typical pay in each sector, so you can read exposure and wages side by side. Click a bar to filter the page to that sector.
How occupations spread across wage bands. The overlay reflects your current filter, so a sector's pay profile stands out against the full set.
Distribution of GAIA-E Score
All 923 occupations · index points (0–100)
Source: GAIA-E composite (gaia_e ×100), Feb 2026 AEI release. Shaded band = interquartile range.
Mean GAIA-E by occupation group
Major O*NET occupation groups · sorted
Source: GAIA-E composite (gaia_e ×100), Feb 2026 AEI release.
Which parts of the workforce are most exposed — and where the workers and wages actually sit. Employment-weighted using US BLS data.
Sector map: exposure × wage × employment
Bubble size = US employment · quadrant lines at the medians
Source: GAIA-E (Eloundou 2024) × BLS OEWS employment & median wage, by major occupation group.
Exposure by occupation group
Employment-weighted mean GAIA-E (index points)
Source: GAIA-E weighted by BLS OEWS employment.
Who bears AI exposure? Occupation exposure crossed with US workforce demographics (American Community Survey / IPUMS), employment-weighted.
What predicts AI exposure?
Correlation of GAIA-E with each demographic (teal = more exposed, coral = less)
Source: GAIA-E (Eloundou 2024) × ACS/IPUMS occupation demographics. Pearson r across 898 occupations.
Exposure by group
Employment-weighted mean GAIA-E, by demographic split
Source: employment-weighted (BLS OEWS) mean exposure by ACS/IPUMS demographic split.
| SOC Code ↕ | Occupation ↕ | Sector ↕ | GAIA-E ↓ | GAIA-B ↕ | Alpha ↕ | Beta ↕ | Gamma ↕ | Median Wage ↕ |
|---|
Compare occupations
All measures side by side
E1 / E1+E2 = % (Eloundou 2024) · GAIA-B = SML score (Brynjolfsson 2018) · GAIA-E = index points · Anthropic observed = AI autonomy (Claude.ai)