$168 Million to Make Every State AI-Ready: Inside NSF's TechAccess Program and Who Should Apply
April 7, 2026 · 7 min read
Arthur Griffin
The National Science Foundation just announced the largest federal investment in civilian AI workforce readiness in U.S. history — and most of the organizations best positioned to win have never applied for an NSF grant before.
TechAccess: AI-Ready America will fund up to 56 Coordination Hubs — one in every U.S. state, the District of Columbia, and every territory — with up to $1 million per year for three years. At full scale, that's $168 million flowing to universities, workforce development boards, community organizations, and state agencies to build the infrastructure that connects American workers, businesses, and governments to AI tools and training. Letters of intent for the first round of 10 hubs are due June 16, with full proposals due July 16.
The program launched on March 27 with a rare multi-agency announcement. NSF, the Department of Labor, USDA's National Institute of Food and Agriculture, and the Small Business Administration are co-sponsoring the initiative — a signal that AI readiness has become a cross-cutting federal priority that transcends any single agency's mission.
"NSF AI-Ready America provides that foundation — giving workers, businesses, and communities in every state and territory the tools and knowledge to advance AI together," said Brian Stone, NSF's point person for the initiative. Labor Secretary Lori Chavez-DeRemer was more direct: "The AI-Ready America initiative will ensure every American worker has the skills, knowledge, and training needed to succeed in an AI-driven economy."
How the Program Actually Works
TechAccess is structured as three interlocking components, each funded through separate solicitations:
State/Territory Coordination Hubs are the core. Each hub serves as the central node for AI readiness efforts within its state, connecting existing programs, identifying gaps, and coordinating deployment of AI tools and training across education, workforce, business, and government sectors. Think of it as a state-level AI extension service — modeled loosely on the land-grant university cooperative extension system that transformed American agriculture.
The 56 hubs will be selected across three competition rounds: 10 in Round 1 (proposals due July 16, 2026), 20 in Round 2 (January 15, 2027), and the remaining 26 in Round 3 (July 1, 2027). This staged approach lets NSF learn from early hubs before scaling, and gives later applicants time to study what worked.
A National Coordination Lead — not yet solicited — will eventually facilitate collaboration between hubs, maintain national dashboards and best-practice repositories, and coordinate sector-specific initiatives. This role will likely go to a major research university or national nonprofit with existing multi-state networks.
AI-Ready Catalyst Award Competitions will fund innovative pilot projects that emerge from hub activities. These haven't been announced yet but represent additional downstream funding for organizations connected to winning hubs.
What Hubs Are Required to Do
NSF's solicitation (NSF 26-508) defines five core responsibilities for each Coordination Hub, and understanding them is essential to building a competitive proposal.
AI Learning Navigator. Every hub must create and maintain "a publicly accessible and user-friendly inventory of state and territory AI-related resources." This means mapping every AI training program, educational pathway, business support service, and government initiative in your state into a searchable, maintained database. Organizations with existing asset-mapping capabilities — workforce investment boards, state education agencies, university extension programs — have a structural head start.
Strategic Planning. Hubs develop statewide AI readiness plans with integrated data collection and evaluation frameworks. This is not a one-time plan; it's an ongoing strategic process that identifies gaps in AI access, sets priorities, and measures progress. The requirement for "integrated data collection" means hubs need systems to track training completion rates, business adoption metrics, and workforce outcomes.
Deployment Support. Hubs provide hands-on assistance for AI adoption to businesses, government entities, and organizations. This is where the extension service analogy becomes concrete — hub staff go to businesses and local governments and help them implement AI tools. NSF envisions an "AI Deployment Corps" of trained specialists who provide direct technical assistance.
Training and Capacity Building. Hubs coordinate K-16 and workforce partners, expand experiential learning opportunities including internships and apprenticeships, and align with national AI literacy frameworks. This is the traditional workforce development piece, but with an important distinction: NSF wants hubs to connect training to actual deployment, not just produce certificates.
Sector Coordination. Hubs convene stakeholders in priority economic sectors — agriculture, manufacturing, healthcare, public services — to identify sector-specific AI needs and opportunities. This is where USDA NIFA, the Department of Labor, and SBA intersect with NSF's mission.
Who Should Apply — And the One-Per-Institution Rule
Each institution is limited to one proposal, which makes institutional positioning critical. The strongest applicants will be organizations with demonstrated statewide convening capacity — the ability to pull together K-12 educators, community colleges, four-year universities, workforce boards, business associations, and government agencies into a functioning coordination network.
Land-grant universities are the most natural fit. They already operate statewide extension networks, have relationships with rural communities, maintain workforce development partnerships, and house AI research programs. In states where the flagship land-grant has a cooperative extension presence in every county, the infrastructure alignment is almost perfect.
State workforce development boards and state education agencies have the convening authority and data systems but may lack the AI technical expertise. A workforce board leading a consortium with university AI researchers and community college training programs could be compelling.
Community college systems with statewide reach — particularly those that have already launched AI training programs — can position around the workforce pipeline angle. The experiential learning requirement (internships, apprenticeships, project-based work) plays directly to community college strengths.
Tribal colleges and universities in states with significant Native populations should consider applying, particularly given USDA NIFA's involvement. AI adoption in agriculture and rural communities intersects directly with tribal economic development priorities.
State-level nonprofits and associations with existing technology access missions could apply, though they'd need to demonstrate the same convening capacity and technical depth as institutional applicants.
The one-proposal-per-institution limit creates an important dynamic: within large universities, different departments or centers may compete internally for the right to submit the institutional proposal. AI research centers, workforce development offices, extension services, and education schools all have legitimate claims. Institutions that resolve this internal competition early — and build genuinely cross-departmental proposals — will produce stronger applications.
The Evaluation Framework
Beyond standard NSF merit review criteria (Intellectual Merit and Broader Impacts), reviewers will assess five program-specific dimensions:
- Vision alignment — how clearly the proposal articulates a statewide AI readiness strategy that maps to NSF's program goals
- Convening capacity — whether the lead organization and partners can demonstrably coordinate across the state's diverse stakeholders
- Gap analysis — quality of the applicant's assessment of current AI efforts and strategy for addressing identified gaps
- Measurable outcomes — realistic milestones with specific metrics: individuals trained by category, small businesses assisted, government entities reached, convenings held, Deployment Corps members activated
- Resource mobilization — credible strategies for securing additional funding beyond the $3 million NSF award
That last criterion deserves attention. NSF explicitly prohibits "voluntary committed cost sharing" — you can't pledge matching funds — but evaluators will assess whether your hub can attract supplementary resources. Organizations embedded in state economic development ecosystems, with relationships to governors' offices, state legislatures, and regional foundations, can credibly argue they'll leverage NSF funding into a larger investment.
The Informational Webinar — Don't Miss It
NSF is hosting an informational webinar on Tuesday, April 14 at 1 p.m. EDT. For a program of this scale and novelty, the webinar is likely to provide critical guidance on what NSF program officers are looking for — the kind of informal signals that don't appear in the written solicitation but shape reviewer expectations. Organizations considering an application should have their proposal lead and key partners attend.
Why This Program Is Strategically Different
Most federal AI investments flow through defense agencies (DARPA, DoD) or research agencies (NSF's core research directorates, DOE national labs). TechAccess breaks that pattern by focusing on adoption rather than invention. The program doesn't fund AI research. It funds the infrastructure that helps ordinary workers, Main Street businesses, county governments, and community organizations use AI tools that already exist.
This distinction matters for grant seekers because it shifts the competitive advantage. The strongest applicants won't be the institutions with the most AI publications or the biggest GPU clusters. They'll be the ones that can demonstrate they know how to meet a farmer in her field, a small manufacturer on his shop floor, or a county clerk at her desk — and help them integrate AI into work they're already doing.
SBA Administrator Kelly Loeffler framed it in competitive terms: "Winning the AI race is vital to ensuring continued American dominance in defense, innovation, and economic strength." But USDA NIFA Director Jaye Hamby offered the ground-level version: "By investing in tools and training that meet farmers and ranchers where they are, we're helping build an agricultural future that is more resilient, more efficient and more accessible."
The tension between those two framings — geopolitical competition versus practical community benefit — runs through the entire program. Smart applicants will address both.
Timeline and Next Steps
The clock is already running. Round 1 selects only 10 hubs, meaning competition will be intense for the initial cohort. But states not selected in Round 1 aren't eliminated — they can reapply in Rounds 2 and 3 with strengthened proposals that incorporate lessons from early winners.
April 14, 2026: Informational webinar (1 p.m. EDT) June 16, 2026: Letters of Intent due (Round 1) July 16, 2026: Full proposals due (Round 1) January 15, 2027: Full proposals due (Round 2, 20 hubs) July 1, 2027: Full proposals due (Round 3, remaining hubs)
Organizations that start building their statewide consortiums now — before the LOI deadline — will have a decisive advantage over those who wait for the webinar to begin outreach. The partnerships section of the proposal requires letters of collaboration from all partners, and those letters take time to secure when you're asking a dozen organizations across different sectors to commit to a three-year coordination effort.
For states that have never had a centralized AI readiness strategy, TechAccess is an invitation to build one with federal backing — and Granted can help you map the competitive landscape and identify the partners that strengthen your consortium before the June deadline arrives.