The Two-State Problem

America is not entering the AI economy as one country. It is entering as two.

One economy is concentrated in a handful of metropolitan corridors where AI capital, talent, research, infrastructure, and policy influence reinforce one another. The other is made up of rural towns, small cities, tribal nations, agricultural communities, and resource regions that sit largely outside the institutions now shaping AI's future. The gap between them is not mainly about who can use a chatbot. It is about who is present in the systems that govern how AI is trained, deployed, procured, financed, and commercialized.

Broadband matters, but connectivity alone won't provide the training, data rights, institutional capacity, sector-specific use cases, or market access that rural communities need to shape and benefit from AI. The real risk is absence from the decisions, not exclusion from the tools.

That would be a national mistake, because rural America is foundational to the country's food, energy, natural resources, logistics, and manufacturing. If AI systems are built without rural context, rural data, and rural use cases, they will perform worse in precisely the environments where the country most needs resilience and trust.

The stakes are large. McKinsey estimates that generative AI alone could deliver $2.6 to $4.4 trillion in annual economic value, and the early evidence suggests those gains will not be distributed evenly.

The Geography of AI Concentration

The San Francisco Bay Area alone accounts for roughly 13 percent of all U.S. AI job postings while representing only a small share of national employment. The top 30 metropolitan areas capture about two-thirds of the country's AI job postings. According to the Brookings Institution, more than half of U.S. metro areas still rank in the bottom two tiers of AI readiness — and those are the metros. Rural America barely registers in most AI-economy metrics at all.

The investment picture is even more lopsided. Global AI venture capital reached extraordinary levels in 2025, with a disproportionate share flowing to the Bay Area, even as the Center on Rural Innovation estimates that less than 2% of all venture capital reaches rural businesses.

None of this is new. NBER research on technology diffusion shows that major technologies tend to originate in a narrow set of places and can take decades to spread. Left to market forces, AI will concentrate where the capital, talent, universities, compute, and customers already are, reinforcing the same geographic inequality that has defined high-wage technology employment for decades.

Why AI Is Different

Rural America has been on the wrong side of technology waves before. But AI is different in four important ways. First, it changes the economics of expertise — AI can give small organizations capabilities that once required consultants or full-time staff. Second, it raises the value of data — rural America generates enormous real-world data in agriculture, health, energy, water, wildfire, logistics, and land use. Third, it reshapes workforce competition. Fourth, it will increasingly govern access to essential services: healthcare triage, lending, insurance, public benefits, education, and local government all run on AI systems, and if rural realities aren't represented, urban assumptions get embedded in tools used everywhere.

The Baseline Divide

Roughly 60 million Americans live in rural communities across most of the nation's land area (using the broader U.S. Census definition; CRAI's default figure is 46 million, ~14%, per the USDA Economic Research Service). Their economic contribution is substantial, yet by nearly every measure of opportunity, rural America has been losing ground for decades. What makes AI uniquely dangerous here is its speed: communities that miss the early wave will miss the period when use cases are defined, data standards are set, workforce pathways are built, and early adopters lock in their advantage.

Colorado as the Proof of Concept

Colorado has built a real AI ecosystem: federal research labs, leading universities, major corporate AI operations, and active policy engagement. But almost all of it sits along the Denver–Boulder–Fort Collins corridor — roughly 70 by 25 miles. Fort Lewis College's AI Institute in Durango stands out as one of the few exceptions: a regional institution serving the Four Corners and directly relevant to rural, Indigenous, and place-based AI participation. The Eastern Plains, the San Luis Valley, the Four Corners, the Western Slope, and the mountain communities face overlapping broadband gaps, limited training infrastructure, and no statewide mechanism to connect them to the AI economy being built elsewhere in the state.

The state's two federally recognized tribal nations make the point. The Southern Ute Indian Tribe has invested heavily in broadband, including a tribally owned open-access fiber network, and the Ute Mountain Ute Tribe has received federal funding for middle-mile fiber. These investments are substantial, but they don't yet answer the harder questions: who governs tribal data, who benefits from AI trained on local knowledge, and how Indigenous sovereignty is protected as AI enters healthcare, education, land management, and public administration.

History Tells Two Stories

In 1935, only about 10 percent of rural homes had electricity; private utilities had decided rural service was unprofitable. The Rural Electrification Administration changed the economics through cooperative financing and deliberate institutional design, and within roughly 25 years nearly every farm in America was electrified. Broadband is the cautionary tale: despite more than $100 billion in cumulative federal investment, the rural-urban gap has persisted for nearly three decades because connectivity alone never generated the jobs, training, institutions, or business models rural communities need. AI requires both lessons at once — the scale and urgency of rural electrification, plus the institutional sophistication broadband lacked.

The Decision Point

America built railroads that connected the coasts, electrified the farms that fed the cities, and laid the cables that eventually reached most of the country. Each required a deliberate decision that the market alone would not solve the problem. AI is no different — except that it moves faster, and the cost of delay is higher. The question is whether rural communities help shape it as producers, data stewards, entrepreneurs, workers, and governing partners — or inherit systems built elsewhere. The two-state problem is solvable, but it won't be solved by broadband alone, by market forces alone, or by urban AI strategies with rural footnotes. The time to build the alternative is now.

Andrew Aitken is the Founder and Executive Director of the Center for Rural AI (ruralai.org), a fiscally sponsored project of SW Community Foundation based in Durango, Colorado. CRAI is partnered with the AI Institute at Fort Lewis College.

Sources: Brookings Metro, Mapping the AI Economy; McKinsey Global Institute; USDA Economic Research Service; Stanford HAI AI Index; Colorado Landscape AI Report; Center on Rural Innovation; Chartis Rural Health Report; Center for Rural AI Rural AI Pilot Program.