The 2026 Edition · Employment data 2025 · Compiled by Christopher Martin

Colorado AI Exposure Atlas

The one-page summary — set to print on a single sheet.

About 1 in 4 jobs in Colorado — roughly 784,400 of 2,860,276, or 27.4% — are in occupations where most or a large minority of tasks overlap with what AI systems can already do. Nationally the figure is 25.9%. Measured by payroll rather than headcount Colorado's share rises to 34.8%, because the exposed work is concentrated in higher-paid office occupations along the Front Range. Across all Colorado jobs, 6.3% are in high-overlap occupations, 21.1% substantial, 27.9% some, and 44.7% little.

High + substantial share of jobs
27.4% about 1 in 4 jobs · score ≥ 45
High + substantial jobs
784,367 of 2,860,276 UI-covered jobs
High + substantial share of payroll
34.8% exposure skews toward higher pay
United States
25.9% same method, national employment

The ten counties with the largest share

Ordered by the share of jobs in high- or substantial-overlap occupations — those scoring 45 or above on the 0–100 measure. The band compares a county with the typical (median) Colorado county, not with the statewide figure. All ten fall in the same band.

The ten Colorado counties with the largest share of jobs in high- or substantial-overlap occupations, with each county's high+substantial and UI-covered employment and its exposure band.
County High + substantial share High + substantial jobs UI-covered jobs Band
Broomfield 35.3% 14,367 40,750 Well above typical
Denver 33.1% 185,864 560,749 Well above typical
Boulder 32.6% 62,359 191,474 Well above typical
Arapahoe 31.0% 105,532 340,882 Well above typical
Douglas 30.8% 46,029 149,574 Well above typical
Crowley Heavily imputed — 54.8% of jobs 28.8% 322 1,121 Well above typical
Jefferson 27.1% 67,103 247,212 Well above typical
El Paso 26.4% 81,447 308,056 Well above typical
Bent Heavily imputed — 55.4% of jobs 25.3% 296 1,172 Well above typical
Larimer 24.7% 42,348 171,290 Well above typical

† 10–50% of county employment imputed. ‡ Over half of county employment imputed — treat as a rough estimate. County industry detail suppressed by QCEW is estimated from state-level mix; see the methodology page. Where more than 30% of a county's employment is imputed the atlas withholds that county's rank outright; the band and the marker above stand in its place. No county is numbered 1–64 on this sheet: the ordinal orders the 0–100 mean score, and all 64 counties sit between 24.7 and 35.2 on it, so adjacent ranks separate on differences the estimates do not support. Ranks and the mean score are published with the dataset.

The method in one paragraph

The atlas combines county employment records (BLS QCEW, 2025 annual averages) with national occupation staffing patterns (BLS OEWS) and published occupation-level AI exposure scores (Eloundou et al., Science, 2024). Occupations are sorted into four tiers on that 0–100 measure — high overlap (≥ 60), substantial (≥ 45), some (≥ 30), and little — and a county's headline figure is the share of its jobs in the top two tiers (score 45 or more). The former binary cut at 50 is published with the dataset as an alternative cut. The five bands cut that share into fifths of about 13 counties each, labeled against the typical (median) Colorado county, which sits at 21.2%. The statewide 27.4% is employment-weighted and therefore runs above 58 of the 64 counties, because most Colorado jobs are in a handful of metro counties. County occupation mixes are estimated from industry mix, not observed, and every county page reports how much of its underlying data was imputed. Exposure is not a job-loss forecast: it measures where AI-driven change — augmentation and automation both — is likely to arrive first. Full methods, sources, validation, and limitations: https://coloradoaiexposureatlas.com/methodology/

Colorado AI Exposure Atlas


All 64 counties, full dataset (CC BY 4.0), and methodology:
https://coloradoaiexposureatlas.com/
Cite as: Colorado AI Exposure Atlas, 2026 edition. Christopher Martin. https://coloradoaiexposureatlas.com/

2026 Edition · Employment data 2025 · Compiled by Christopher Martin