Every state graduates students across the full range of academic fields, but not in the same proportions. This map shows the average AI exposure of each state's Academic Program Portfolio, weighting every academic program by the number of credentials awarded. It combines the Opportunity Data AI Exposure Index with 2023 IPEDS completions, covering all 50 states and the District of Columbia.
Most analyses measure the AI exposure of occupations or workers. This analysis measures the AI exposure of the educational programs from which students graduate, providing a view of the educational pipeline rather than the current workforce.
Completion-weighted AI exposure of program graduates, by state of institution. Darker means a higher state AI exposure score; the scale is stretched to the observed 51-state range. 2023 award year, 5.3M completions. Sources: Opportunity Data AI Exposure Index, built from O*NET work characteristics weighted by BLS OEWS employment; completions and institutional directory from NCES IPEDS.
Two things are true at once: state differences in AI exposure are small, and they are real. Because every state graduates students across the full range of fields, state averages compress into a narrow band, from 0.500 in Nevada to 0.588 in the District of Columbia, around a national mean of 0.540. No state's graduates are dominated by AI-exposed fields, and none escape them. The compression is the finding, not a flaw in the measure. Exposure is a feature of fields of study, and every state teaches the full spectrum.
Within that narrow band the ordering is clear and not random. District of Columbia, Maryland, New Hampshire, and Massachusetts sit highest because their graduates concentrate in analytic, business, and professional fields, and above all in advanced credentials: three in five completions in the District of Columbia are at the graduate level. States toward the lighter end graduate proportionally more in trades, personal services, and hands-on health programs, where the work stays anchored to people and places. Exposure here is a property of what a state graduates, not where its graduates end up working, and a higher score is not a worse one. It marks where the question of how a credential holds its value is most live.
These findings are developed state by state, with the most and least exposed states, the credential lever behind the gap, the field-within-level exceptions, and what it means for Workforce Pell, in the findings brief.
Read the findings brief (PDF) →All 51 jurisdictions on one scale. Bars are scaled within the observed range (0.500–0.588) to make the gradient visible, so adjacent states differ by less than the bars suggest. Darker bars are higher scores.
The ranking reflects each state's Academic Program Portfolio for 2023: the fields its graduates earned credentials in, weighted by how many earned each. A state rises when its graduates cluster in analytic and professional programs and falls when they cluster in trades, services, and hands-on health. The credential columns show how that portfolio splits across award levels. Click a column to sort.
| # | State | State AI exposure score | Completions | Cert <1y | Cert 1–2y | Assoc. | Bach. | Grad. |
|---|
Each state's score is the completion-weighted average AI exposure of the programs its graduates complete. Program exposure comes from the AI Exposure Index, a composite of digital intensity, human-interaction shield, and physical anchor scored for 1,786 academic programs; completion counts come from IPEDS for the 2023 award year (first majors, 5.3 million awards) and are mapped to states through the IPEDS institutional directory. Programs the index does not score, about one in eight completions, are held in the denominator at the national average exposure (0.542) so they neither raise nor lower a state's score. The color scale runs light to dark across the observed 51-state range, so differences of a few thousandths are visible; because the true range is narrow, read position relative to other states rather than as a large absolute gap.
The full methodology note documents the index construction, the IPEDS completion weighting, the CIP join, the neutral treatment of unscored programs, the award-level tiers, and the limits of the measure.
Download the methodology note (PDF) →Opportunity Data. (2026). AI Exposure of Academic Programs by State. opportunitydata.org/ai-exposure-states.html
Opportunity Data works with institutions, agencies, and researchers on analysis of education and workforce data. For collaboration, custom analysis, or questions about methods and sources, get in touch. Feedback and suggestions are welcome.
Methods, sources, a state you want broken out further, or a collaboration. Send a note.