NSF's TechAccess Bets $1M a Year on One Organization in Every State — And the First 10 Hubs Are Decided July 16.

Most federal grant competitions ask you to be the best applicant. NSF's new TechAccess: AI-Ready America program (solicitation NSF 26-508) asks something stranger and harder: to be the only applicant your state should have. The program will fund up to 56 State/Territory Coordination Hubs — one per state, the District of Columbia, and each U.S. territory — at $1 million per year for three years, with a possible fourth transition year. That is a total program budget in the range of $168–224 million, and it is structured so that within any given state, the money converges on a single backbone organization.

The first round is nearly here. The full proposal deadline for Round 1 is July 16, 2026, and only 10 hubs will be selected in this cohort. Round 2 (20 hubs, due January 15, 2027) and Round 3 (the remainder, due July 1, 2027) follow, but the states that win early set the template — and the partnerships — that later applicants will have to work around. This is the deep dive on what a hub actually is, who can credibly lead one, and how to write the 15 pages that decide it.

What a Coordination Hub actually does

TechAccess is not a research grant, and reading it as one is the fastest way to lose. NSF built the program to move AI capability out of universities and into the parts of the economy that don't have a computer-science department: small businesses, local governments, community colleges, workforce boards, agricultural extension offices, and public-serving nonprofits. The hub is the connective tissue. NSF's own language calls it a "backbone organization" — a term drawn from collective-impact practice, not from the lab.

The solicitation names five responsibility areas every hub must execute, and a competitive proposal has to show a concrete plan for all five:

1. AI Learning Resource Navigator — build and maintain a public inventory of the AI training programs, infrastructure, and support services that already exist in your state.
2. Strategic Planning — convene partners to write a statewide AI-readiness plan with real data collection and evaluation baked in.
3. AI Deployment Support — provide hands-on help to local governments, small businesses, and community organizations adopting AI, potentially through a credentialed AI Deployment Corps that plugs into a national network.
4. Training and Capacity Building — coordinate K-16 and workforce systems, incorporate micro-credentials and digital badges, and expand apprenticeships and project-based learning. This must align with the DOL AI Literacy Framework and federal workforce law (WIOA and Perkins V).
5. Sector-Specific Coordination — convene priority industries (energy, agriculture, manufacturing, healthcare, education) around shared AI-adoption and training strategies.

Notice what unites these: none of them are things a single institution does alone. Every one is a coordination task. NSF is not funding the organization with the best AI research; it is funding the organization that can get everyone else in the state to show up in the same room and keep showing up for three years.

Who can actually win the backbone role

Eligibility is broad on paper and narrow in practice. Any organization eligible under NSF's PAPPG rules can apply, and each institution may submit only one proposal. Unaffiliated individuals cannot apply. Critically, voluntary committed cost sharing is prohibited — so you cannot buy your way to competitiveness by promising matching dollars, and NSF will not reward proposals that try.

What NSF will reward is demonstrated statewide convening power. The solicitation asks the lead organization to prove it can operate across an entire state and manage multi-sector initiatives. AI expertise is required but — importantly — it "may be provided through partnerships." That single clause reshapes the competition. It means the natural lead is often not the flagship research university's AI lab, but the organization that already sits at the center of the state's economic and workforce networks: a university system office, a state's land-grant Cooperative Extension service, a community college system, a statewide nonprofit intermediary, or a public-private economic-development alliance. Those organizations can source AI expertise from a partner university while bringing the one thing a lab usually can't — reach into all 67 counties.

If you are a research university reading this, the strategic question is not "can we win?" but "should we lead or should we be the AI-expertise partner on someone else's application?" Because only one proposal per institution counts, and only one hub per state gets funded, the wrong choice is to have three units at your university each assume someone else is coordinating the bid. States that field a single, unified, cross-sector proposal will beat states that field three fractured ones.

How the proposal is actually scored

The Project Description is capped at 15 pages and must contain five specific section headers — miss them and NSF returns the proposal without review. The sections map directly onto what reviewers score:

• Section 1 — Vision and Approach: how you'll execute all five responsibility areas, with evidence of prior statewide work.
• Section 2 — Team and Governance: the lead organization's profile, partner rationale, and — this is where many proposals are won or lost — the governance structure that will hold the coalition together.
• Section 3 — State Landscape: an honest map of what's already working in your state, where the gaps are, and how the hub accelerates rather than duplicates existing efforts.
• Section 4 — Work Plan: milestones, timelines, and measurable performance metrics — number of individuals trained, small businesses and government entities assisted, convenings held, Deployment Corps members trained, contributions to national repositories.
• Section 5 — Sustainability: credible strategies for attracting additional resources from industry, philanthropy, and state agencies once NSF money tapers.

Letters of Collaboration from every named partner are mandatory. Letters of support are prohibited — NSF wants commitments, not applause. That distinction is a tell about how the program will be judged: reviewers are looking for a real, pre-existing coalition, not a wish list of organizations you hope to recruit after the award lands.

The strategic reality of a three-round structure

The single most consequential feature of TechAccess is that it runs in three rounds with only one hub per state. If a well-organized coalition in your state wins in Round 1, there is no second bite — your state's hub is decided. That creates a genuine first-mover dynamic that most NSF competitions don't have. For strong, ready coalitions, Round 1's July 16 deadline is not a convenience; it is the whole game. For coalitions that aren't ready, the honest move is to aim at Round 2 (January 15, 2027) and spend the intervening six months building the governance structure and the signed partner letters that Round 1 winners already have.

There is also a mandatory downstream obligation worth planning for now: every hub must actively collaborate with a National Coordination Lead (funded separately), contributing to national dashboards, participating in convenings, and feeding a shared best-practice repository. Build proposals that treat national reporting as a first-class activity, not an afterthought — the metrics in your Section 4 work plan should already be framed to roll up into a national dashboard.

The bottom line

TechAccess: AI-Ready America is one of the most unusual federal opportunities of 2026 — a program where being the strongest applicant matters less than being the right applicant for your state's whole ecosystem. The organizations that win Round 1 on July 16 will be the ones that spent the spring assembling a genuine coalition, securing signed Letters of Collaboration, and writing a governance plan that reviewers can believe will still be standing in year three. If that describes your organization, the next two weeks are the most important of the cycle. If it doesn't yet, Round 2 is your build window — and it starts now.

Granted tracks NSF AI and workforce solicitations as they open. See our complete guide to NSF AI funding programs for the full 2026 calendar.