If your strategic AI plan treats US AI investment as broadly distributed across the country's economic geography, the state-level breakdown shows you are working from a model the concentration data has obsoleted. California alone captured $218 billion in cumulative AI investment over the 2013–2025 period. That is 75% of total US national AI investment. The remaining 49 states combined captured the other 25%.

US state-level AI investment concentration, 2013–25:

The data shape. The United States cumulative AI investment from 2013 to 2025 was $757.27 billion. Of that, California accounted for $218 billion. New York, Massachusetts, Texas, and Washington, the next four states, combined for roughly another $150 billion. The remaining 45 states divided the remaining ~$390 billion among them, with most states capturing less than $5 billion each over the entire 12-year period.

Within California, the concentration is even tighter. The San Francisco Bay Area (including Silicon Valley) captures the majority of California's AI investment. The rest of California (Los Angeles, San Diego, Sacramento) captures meaningful but smaller volumes. Within the Bay Area, the concentration is in a small number of cities: San Francisco, Palo Alto, Mountain View, Menlo Park, Sunnyvale.

The 2025 single-year figures tighten the concentration further. With $285.88 billion in US AI investment in 2025 alone, California captured a disproportionate share even of this single year. The state's share of US AI investment is rising, not falling, even as absolute amounts grow.

AI investment by country, 2013–25:

For strategic and procurement teams, the concentration has practical implications.

The first: vendor proximity matters more than the global investment picture suggests. The frontier AI providers (OpenAI, Anthropic, Google DeepMind, Meta AI, xAI) are headquartered in a roughly 50-mile radius around San Francisco. The talent pool, the partner ecosystem, the venture capital infrastructure, and the technical advisory networks are all concentrated in the same geography. Enterprises located far from this concentration face structural disadvantages in vendor relationships, talent recruitment, and partnership access that the global "AI everywhere" narrative obscures.

The second: alternative geographies are not catching up at the rate the rhetoric suggests. Texas (Austin), New York, Boston, and Seattle have AI investment activity, but the gap to California is growing, not shrinking. Texas's AI investment grew substantially in 2024–2025 (driven by Stargate, the Google Texas data centre, and a few large deals), but California's growth was larger in absolute terms. The argument that "AI is distributing across the US" is true in nominal terms but false relative to California.

The third: the concentration extends beyond capital to talent and infrastructure. California captured the bulk of US AI investment, and the parallel data on AI researchers, AI startups, and AI infrastructure is similarly concentrated. The AI workforce, the AI startup ecosystem, and the AI capital are all in roughly the same geography. This is unusual: most major technology categories distribute these three layers across different geographies.

The prescription for strategic planning needs to be specific about which problem the concentration creates.

For enterprises operating outside California, the concentration creates a procurement and partnership challenge. Vendor relationships built remotely lack the access to senior vendor leadership and forward-roadmap discussions that California-based enterprises receive. Partnerships built remotely lack the informal discovery and trust-building that proximity enables. Talent recruitment outside California faces a thinner pool of AI-experienced workers. Strategic plans should account for these structural disadvantages, not by relocating (impractical for most enterprises) but by investing more heavily in compensating mechanisms: senior vendor relationship management, structured partnership development programmes, remote-first AI talent recruitment, and direct presence in California for executives and senior technical staff.

For governments and economic development agencies outside California, the concentration creates a policy challenge. The diffusion of AI capability from California to other geographies will not happen automatically. The factors that built California's concentration (venture capital infrastructure, academic research density, immigration policies, and historical accident) are difficult to replicate quickly. Policy efforts to build AI ecosystems elsewhere need to be sized against the structural advantages California has accumulated over decades. Modest investment will not move the needle; substantial sustained investment over 10–20 years might.

For California-based enterprises, the concentration creates a different problem: workforce cost inflation, real estate cost inflation, and the talent retention challenge of competing with frontier-model providers paying compensation that few enterprises can match. The benefits of being in the concentration are real (access, partnerships, talent pool), but the costs are also rising. Strategic plans for California-based AI activity should account for cost trajectories that are higher than national averages.

The trajectory through 2026–2028: the concentration will continue, possibly tightening further. The structural forces that drove the concentration (capital availability, talent agglomeration, partnership ecosystem, research density) are reinforcing rather than dispersing. Strategic plans built for the concentration that the data describes will produce better outcomes than plans built on the assumption that AI investment is broadly distributed.

The prescription: design AI strategy specifically for the concentration. Enterprises outside California should invest in compensating mechanisms (talent, partnership, vendor relationship). Enterprises in California should plan for cost trajectories and competitive talent dynamics. Governments and economic development agencies should size policy efforts to the scale of the structural advantage they are working against. The "AI is everywhere" framing is rhetorically attractive but does not match the data. Strategic plans built on the data will outperform plans built on the rhetoric.

For executive teams setting AI direction, the planning anchor should explicitly identify whether the enterprise is operating from inside or outside the concentration, and design strategy accordingly. The two situations are structurally different. Strategy that ignores the difference will produce worse outcomes than strategy that addresses it. By 2028, the structural advantages California has accumulated will be wider, not narrower, and the strategic question of inside or outside the concentration will have higher consequence than it does today.

Sources

  • Primary: Stanford AI Index 2026, Chapter 4 (Economy) 4.2 — hai.stanford.edu/ai-index/2026
  • Investment data: Quid, 2025 — US AI investment by state 2013–25; California concentration breakdown
  • Cross-country investment: Quid global AI investment dataset, 2013–25