NSF Is Spending $224 Million to Put an AI Hub in Every State. Here's How to Win One.

The National Science Foundation just announced the largest AI workforce and literacy program in its history: $224 million to establish coordination hubs in all 50 states, the District of Columbia, and five U.S. territories. Each hub gets up to $1 million per year for three years. Letters of intent for the first round of ten hubs are due June 16, 2026, with full proposals due July 16.

This isn't a research grant. It's an infrastructure play — a federally funded effort to ensure that every American worker, small business owner, local government employee, and K-16 student has access to AI literacy, tools, and training. The program is called TechAccess: AI-Ready America, and it represents a bet that AI readiness is a coordination problem, not just an education problem. The states that organize best will win first.

Why This Program Exists

The gap that TechAccess targets is real and measurable. AI adoption in the United States is concentrated in a handful of metropolitan corridors — the Bay Area, Boston-Cambridge, Seattle, Austin, New York. Outside those clusters, most small businesses have no systematic access to AI training. Most community colleges don't offer applied AI coursework. Most state workforce development systems haven't integrated AI skills into their credentialing frameworks. Most rural school districts can't afford AI literacy programs even when curricula exist.

Federal investment in AI has historically flowed through research grants — NSF's AI institutes, DARPA programs, DOE national lab projects. That money produces cutting-edge capabilities but doesn't solve the distribution problem. A farmer in Nebraska, a county clerk in Mississippi, a machine shop owner in West Virginia — these aren't people who need frontier AI research. They need someone to show them which tools exist, how to use them safely, and where to get help when something goes wrong.

TechAccess is designed to be that someone, in every state simultaneously. The program's multi-agency structure — NSF partnering with the Department of Labor, USDA's National Institute of Food and Agriculture, and the Small Business Administration — reflects the recognition that AI readiness crosses every traditional agency boundary.

What the Hubs Actually Have to Do

The solicitation (NSF 26-508) lays out five core responsibilities for each coordination hub. Understanding these is essential for anyone preparing a proposal, because the evaluation criteria map directly to them.

AI Learning Navigator. Each hub must build and maintain an accessible inventory of AI resources available in its state — existing training programs, university courses, community college certificates, employer-led initiatives, online platforms, and government services. The inventory isn't just a list; the hub must actively curate it, identify gaps, and connect people to appropriate resources based on their starting point and goals. Think of it as a statewide 211 system for AI, except it doesn't exist yet and the hub has to build it.

Strategic Planning. Hubs must develop a statewide AI readiness plan with measurable goals, evaluation mechanisms, and stakeholder buy-in from across sectors. This is where the program gets political in the best sense — the hub has to convene business leaders, educators, workforce boards, tribal nations, rural communities, and state agencies around a shared plan. States that already have AI task forces or governor-level AI initiatives have a head start. States that don't will need to build coalitions from scratch within the proposal timeline.

Deployment Support. The hub provides hands-on technical assistance to organizations trying to adopt AI — not selling software, but helping a local government understand which permitting processes could benefit from automation, or helping a manufacturing firm evaluate whether predictive maintenance tools make sense for their equipment. This is the extension-service model applied to AI: trusted advisors who understand local context.

Training and Capacity Building. Hubs coordinate AI education across the K-16 pipeline and adult workforce systems. This means working with state departments of education on K-12 AI literacy standards, with community colleges on certificate programs, with universities on applied AI curricula, and with workforce boards on integrating AI competencies into registered apprenticeship programs. The DOL partnership is explicitly designed to connect hubs to the apprenticeship infrastructure that DOL's recent $85 million State Apprenticeship Expansion grants are building out.

Sector Coordination. Each hub must convene stakeholders in priority economic sectors — agriculture, healthcare, manufacturing, government services, education — to identify sector-specific AI adoption barriers and coordinate solutions. The USDA NIFA partnership targets agricultural communities specifically, while SBA involvement ensures small business needs are represented.

Who Should Apply

NSF limits submissions to one proposal per organization, and the program is clearly designed for institutions with statewide convening power. The strongest applicants will be:

Land-grant universities are natural fits, particularly those with established cooperative extension networks. The extension model — trusted advisors embedded in communities, translating research into practical application — is essentially what TechAccess is trying to build for AI. Universities with extension offices in every county already have the physical infrastructure and community relationships that a hub needs.

State workforce development agencies or their affiliated organizations, particularly those with existing relationships to the DOL apprenticeship system. If your state's workforce board already coordinates with community colleges and employer networks, you're positioned to add AI coordination to an existing architecture.

Community college systems with statewide reach. In states where a single community college system serves the entire state (like Virginia's VCCS or North Carolina's NCCCS), the system office can propose a hub that leverages every campus as a deployment node.

Tribal colleges and universities in states with significant Native American populations can propose hubs that center Indigenous communities' AI readiness needs — an angle that reviewers will value given NSF's broadening participation mandate.

Nonprofit organizations with demonstrated statewide convening capacity, particularly those already working on digital equity or workforce development.

The critical requirement is that the lead organization must demonstrate it can convene stakeholders across sectors and geographies within its state. A university with strong AI research but no extension network or community partnerships will score poorly. A community organization with deep local ties but no statewide reach won't meet the scope. The sweet spot is institutional breadth combined with community credibility.

The Three-Round Strategy

NSF is awarding hubs in three rounds: 10 in Round 1 (proposals due July 2026), 20 in Round 2 (proposals due January 2027), and the remaining 26 in Round 3 (proposals due July 2027). This structure creates both opportunity and strategic considerations.

Round 1 favors states with existing infrastructure. States that already have governor-level AI initiatives, active AI task forces, or university-led AI literacy programs can point to existing momentum. If your state has already conducted an AI workforce needs assessment, that data becomes the foundation of your strategic plan. The 10 Round 1 selections will likely skew toward states that can demonstrate they're building on something rather than starting from zero.

Round 2 is the volume round. With 20 slots, this is where most competitive states will land. The six-month gap between Round 1 and Round 2 gives applicants time to study what won in Round 1 and calibrate. States that applied in Round 1 and weren't selected can resubmit with refined proposals.

Round 3 catches everyone else. By this point, the national coordination lead — a separate award that NSF will make to facilitate collaboration across all hubs — will be active, and later hubs will benefit from templates, shared resources, and lessons learned from earlier cohorts. The tradeoff is that Round 3 hubs have less time to establish themselves before the three-year clock starts ticking.

For states with strong existing capacity, applying in Round 1 is the right move. For states where coalition-building is still underway, Round 2 may be more realistic — and the extra preparation time is worth more than an early start.

What Reviewers Will Look For

The evaluation criteria emphasize five dimensions, and understanding their relative weight matters for proposal construction:

Vision and alignment — Does the proposal articulate a coherent statewide AI readiness strategy that addresses all five hub responsibilities? Generic statements about "democratizing AI" won't score well. Reviewers want to see specific gaps identified in the state's current AI ecosystem and a concrete plan for filling them.

Convening capacity — Can the lead organization actually bring the necessary stakeholders to the table? Letters of support from the governor's office, state workforce board, major employers, community college systems, and tribal nations carry weight. A proposal from a university that hasn't consulted its state's workforce development infrastructure will look disconnected.

Gap analysis — Does the team understand what AI resources already exist in the state and what's missing? The strongest proposals will include a preliminary inventory of existing AI training and deployment resources, identify specific populations or regions that are underserved, and propose targeted interventions for those gaps.

Measurable milestones — The solicitation explicitly asks for realistic milestones and measurable outcomes. "Increase AI literacy" is not measurable. "Train 500 small business owners in applied AI tools across 12 workshops in Year 1, with pre- and post-assessment scores" is.

Sustainability — NSF wants to know what happens after the three-year funding ends. Hubs that articulate a transition plan — state funding, employer contributions, integration into existing workforce systems, fee-for-service models — will score higher than those that assume federal funding continues indefinitely.

The Bigger Picture

TechAccess lands at a moment when the NSF is rebuilding its identity after a bruising year. The agency lost 1,752 grants worth $1.4 billion to DOGE-driven terminations, saw its director resign, and watched its peer review standards diluted. Against that backdrop, TechAccess represents something the agency rarely attempts: a program with a clear deliverable that the public can understand and that the administration actively supports.

The AI workforce angle aligns with stated presidential priorities. Labor Secretary Lori Chavez-DeRemer emphasized ensuring "every American worker has the skills, knowledge, and training needed to succeed in an AI-driven economy." NSF Director Brian Stone framed it as providing "tools and knowledge to advance AI together." That bipartisan alignment gives TechAccess political durability that pure research programs don't enjoy — particularly valuable for a three-year commitment that will span at least one more budget cycle.

For organizations considering whether to invest the effort in a TechAccess proposal, the math is compelling: $3 million over three years to build statewide AI coordination infrastructure, with the institutional relationships and credibility that come from being the federally designated AI hub for your entire state. The letter of intent is four pages. The full proposal is fifteen. The deadline for Round 1 LOIs is June 16 — six weeks from today.

If your organization has statewide reach, a track record in workforce development or technology deployment, and the ability to convene across sectors, this is worth the investment. Tools like Granted can help you scope competing opportunities and build a proposal that meets NSF's evaluation criteria before the deadlines arrive.