Context
Global Expansion
What the data is
Behavioral evidence of AI engagement from public GitHub activity: 5.81M monthly records across 64 months (Jan 2021 – Mar 2026). Every user was caught doing AI work, not surveyed or inferred.
Growth arc
Early base (Jan 2021)
16,323 users
Inflection (Sep 2022)
288K users
Peak activity (Feb 2025)
258K monthly active (record high)
Current (Mar 2026)
2.58M cumulative
Signal Quality
235K users have city-level resolution (9% of total). Most provide no location (86.5% dark), so this is a self-selected, publicly visible slice — conservative floor, not full census.
Structurally shifted
2021: Small specialist group (avg score 95–125). 2025+: Broad range from frontier researchers (score 1000+) to recent adopters on model APIs.
Market
US Concentration
Raw scale
US users (located)
51,913
% of all located
~16%
Estimated true total
~385,000
Top metros (all-time)
- San Francisco — 5,172 (10% of US total)
- New York — 3,309
- Seattle — 2,603
- Los Angeles — 2,420
- Austin — 1,221
Per-capita leaders
DC
572 per M
Washington
361 per M
California
307 per M
Bay Area Dominance
SF-Oakland metro holds 10% of all US AI users while representing <2% of US population. Structural depth, not population-driven.
Momentum
Growth & Action
Fastest-growing metros (2024–2025+)
Raleigh-Durham
+64.7%
Dallas-Fort Worth
+49.3%
Austin
+36.4%
Most intense market
Austin stands out: 2025+ avg score of 519.84 — highest among top-growth metros. Growth + intensity.
US YoY trend
2024 unique users
852K
2025 unique users
1.30M
YoY change
+52.5%
Talent playbook
- Scale markets (SF, Seattle, NYC): Use for competitor sensing and volume hiring.
- Growth markets (Austin, Atlanta, Denver): Build early pipeline before req pressure.
- Momentum plays (Raleigh, Dallas, Portland): First-mover sourcing advantage.