NSF Wants an AI Hub in Every State. The $168 Million Plan to Make It Happen.

Somewhere between the trillion-dollar AI infrastructure announcements and the frontier model arms race, a quieter question has gone largely unanswered: what happens to the 150 million American workers, the 33 million small businesses, and the thousands of local governments that need to use AI but have no idea where to start?

The National Science Foundation thinks the answer is local. On March 25, NSF announced TechAccess: AI-Ready America — a multi-agency initiative to establish coordination hubs in all 50 states, five territories, and the District of Columbia. Each hub will receive up to $1 million per year for three years, with a possible fourth year for those that demonstrate continued need. The total investment: up to $168 million across 56 hubs, making it the largest federal AI workforce and adoption program ever attempted at the state level.

The initiative is not just an NSF effort. The Department of Agriculture's National Institute of Food and Agriculture, the Department of Labor, and the Small Business Administration are co-sponsors — a four-agency partnership that signals the administration views AI readiness as an economic competitiveness issue, not merely a research question.

First-round applications are due July 16, 2026. An informational webinar is scheduled for April 14. And the design of the program reveals a set of assumptions about AI adoption that differ markedly from how the federal government has historically approached technology transfer.

Why State-Level Hubs, and Why Now

The federal government has funded AI research for decades. NSF's National AI Research Institutes — a network of 25 centers across the country — have produced foundational work in areas from agricultural AI to trustworthy computing. DOE's national laboratories run some of the world's most powerful AI systems. DARPA's investments helped create the modern deep learning revolution.

But almost none of this investment has been designed to help a 15-person manufacturing company in rural Ohio adopt AI for quality control, or to train county government employees in Iowa to use AI for permit processing, or to give a community college in Mississippi the curriculum to prepare students for AI-adjacent jobs. The gap between frontier AI research and practical AI adoption at the community level is enormous — and it is the gap TechAccess is designed to close.

NSF Director Brian Stone framed the strategic logic plainly: "America's AI competitiveness depends on research paired with advanced knowledge access for our current and future workforce." SBA Administrator Kelly Loeffler was more direct: "Winning the AI race ensures continued American dominance in defense, innovation, and economic strength."

The state-level hub model borrows from a structure the federal government has used successfully in other contexts. The Manufacturing Extension Partnership (MEP), operated by NIST, places centers in every state to help small manufacturers adopt new technologies. Cooperative Extension, run through USDA and land-grant universities, has delivered agricultural knowledge to rural communities for over a century. TechAccess adapts this hub-and-spoke model for AI, with the explicit recognition that AI adoption requires local context — the needs of a fishing community in Alaska differ from those of a tech corridor in Northern Virginia.

What the Hubs Will Actually Do

The funding opportunity defines three priority areas that each hub must address:

AI literacy and workforce skills. Hubs must expand AI knowledge and applied skills across their state's workforce. This is not limited to technical training — it includes helping workers in non-technical roles understand how AI tools can augment their work, identifying which occupations face the greatest disruption, and building training pathways that connect to actual employment. The Department of Labor's involvement signals that hubs will be expected to coordinate with state workforce development boards, community colleges, and employer networks.

Small business and local government adoption. Hubs must equip small businesses and local governments with AI adoption tools and technical assistance. This is where USDA NIFA's partnership matters most. NIFA Director Jaye Hamby noted that "every community, including rural areas, can benefit from AI through tools and training meeting farmers where they are." For rural economies — where small businesses and local governments often lack dedicated technology staff — the hub model provides a point of contact that does not require flying to a conference in San Francisco or hiring a consultant from Boston.

Hands-on learning pathways. Hubs must build internships, apprenticeships, and project-based learning programs that give workers practical AI experience. Labor Secretary Lori Chavez-DeRemer emphasized that "every American worker needs the skills and training to succeed in an AI-driven economy." The learning pathways are designed to complement, not replace, existing postsecondary programs — connecting four-year universities, community colleges, vocational programs, and employer-based training into a coordinated system.

The Competition Structure

The 56 hubs will be selected across three rounds:

• Round 1: 10 hubs, applications due July 16, 2026
• Round 2: 20 hubs, applications due January 15, 2027
• Round 3: Remaining hubs, applications due July 1, 2027

The staggered timeline is strategic. Round 1 hubs will serve as proof-of-concept models that inform the design and expectations for subsequent rounds. Organizations applying in Round 2 and 3 will have the advantage of seeing what worked and what did not in the first cohort.

NSF also plans to select a national coordination lead — a separate award to a single organization that will facilitate collaboration and knowledge sharing across all 56 hubs. This role has not yet been solicited but is expected to be announced later in 2026. Additionally, NSF anticipates issuing AI-Ready Catalyst award competitions focused on specific topics to pilot and scale innovative approaches that address critical AI readiness needs identified by the hubs.

Who Should Apply

The eligibility for coordination hubs extends to institutions of higher education, nonprofit organizations, and state and local government entities. But the competitive applicants will share certain characteristics.

Land-grant universities and community college systems have a natural advantage. They already have statewide networks, workforce development infrastructure, and relationships with employers and local governments. A land-grant university that partners with its state's community college system and workforce development board can present a compelling case for reaching every corner of the state.

Economic development organizations and regional planning commissions that serve multi-county areas can aggregate demand from small businesses and local governments in ways that individual municipalities cannot. If your organization already provides technical assistance to small businesses — whether through an SBDC, an MEP center, or a regional economic development district — the TechAccess hub model is designed to build on exactly that kind of existing infrastructure.

Tribal colleges and organizations serving underserved communities should note that the program explicitly targets equitable access. Rural, tribal, and economically distressed regions are not afterthoughts in this funding opportunity — they are central to the rationale. Proposals that demonstrate how AI readiness can address specific local economic challenges in underserved areas will be competitive.

Organizations that lack statewide reach or existing workforce development partnerships will find it difficult to compete. NSF is not looking for standalone AI training programs. It is looking for coordination capacity — the ability to convene partners, align resources, and scale proven approaches across an entire state or territory.

The Bigger Picture: Federal AI Investment Beyond Research

TechAccess represents a philosophical shift in how the federal government thinks about AI funding. For years, federal AI dollars flowed overwhelmingly to two destinations: frontier research at elite universities and national labs, and defense applications through DARPA and the military services. The assumption was that breakthroughs at the top would eventually trickle down to the broader economy.

TechAccess inverts that assumption. It says that AI's economic impact depends not on making models more capable — the private sector is spending hundreds of billions on that — but on making the existing workforce, businesses, and institutions capable of using what already exists. The $168 million investment is modest compared to DOE's $293 million Genesis Mission funding or the billions flowing into AI research. But its potential reach — every state, every territory, every small business and local government that walks through a hub's door — is broader than any other federal AI program to date.

For organizations that see themselves as the right hub for their state, the July 16 deadline for Round 1 is less than four months away. Assembling the partnerships, mapping the state's AI readiness gaps, and designing a credible three-year plan takes time — and the strongest applications will demonstrate that the work has already started. Granted can help you identify complementary federal programs, track co-funding opportunities, and build the kind of multi-agency proposal that this four-partner initiative expects.