NSF Announces $100M for AI Research Institutes — What It Means for You

Five new institutes. Twenty-nine total across the country. And $100 million in fresh public-private funding announced in a single day. The National Science Foundation's latest round of AI Research Institute awards, unveiled on July 29, 2025, represents the largest expansion of the program since its 2020 launch -- and a clear signal about where federal AI dollars are headed next.

Browse our AI Grants page for current opportunities.

What NSF Just Funded

The five new institutes, backed jointly by NSF, Capital One, and Intel, each receive approximately $4 million per year over four to five years. They span materials science, machine learning foundations, education, drug discovery, and human-AI interaction:

NSF AI-Materials Institute (AI-MI), led by Cornell University, targets next-generation materials for energy, sustainability, and quantum technologies. The institute will use AI to accelerate the notoriously slow cycle of materials discovery -- a bottleneck that affects everything from battery design to semiconductor manufacturing.

NSF Institute for Foundations of Machine Learning (IFML), led by the University of Texas at Austin, focuses on expanding generative AI into new domains. Their agenda includes diffusion models, protein engineering, and clinical imaging -- areas where foundational ML research can produce outsized practical impact.

NSF Institute for Student AI-Teaming (iSAT), led by the University of Colorado Boulder, is building AI tools that enhance group learning in STEM classrooms. The institute already serves over 6,000 students and is co-developing semester-long curricula to build AI literacy from the ground up.

NSF Molecule Maker Lab Institute (MMLI), led by the University of Illinois Urbana-Champaign, develops AI-driven workflows for molecular discovery -- think autonomous synthesis of new drugs, catalysts, and clean-energy materials. The long-term goal is fully autonomous molecular design.

NSF AI Research Institute on Interaction for AI Assistants (ARIA), led by Brown University, tackles the safety and adaptability of next-generation AI assistants. Capital One is contributing $1 million per year over five years specifically to support ARIA's research.

A sixth award went to UC Davis to operate the NSF AI Institutes Virtual Organization (AIVO), a national hub that connects all 29 federally funded AI Institutes into a coordinated network.

The Bigger Picture: 29 Institutes and Growing

This latest cohort brings the NSF AI Research Institutes network to 29 active programs connecting over 500 institutions worldwide. The program has grown steadily since 2020: seven institutes in the first cohort, eleven added in 2021, seven more in 2023, and two NSF-Simons astronomy institutes announced in September 2024, each receiving $20 million in joint funding.

The themes tell the story of evolving national priorities. Early rounds emphasized agriculture, weather forecasting, and cybersecurity. The 2023 cohort added astronomy and materials research. Now the program is pushing into AI safety, autonomous scientific discovery, and education at scale. Each generation of awards reflects what policymakers see as the next critical frontier -- and where they expect to write the next round of checks.

The program also aligns with the White House AI Action Plan and Executive Order 14277 on AI education, giving these institutes political tailwinds that could sustain funding even in tight budget years.

Who Can Apply (and What to Watch For)

Under the most recent solicitation (NSF 23-610), eligibility extends to two-year and four-year institutions of higher education as well as nonprofit, non-academic organizations. Awards run $16 million to $20 million over four to five years. Institutions may submit a maximum of two proposals, and any individual can serve as PI or co-PI on only one.

The NSF AI Research Institutes program page currently shows no upcoming due dates -- the program is waiting for a new solicitation to be published. That is the critical window to watch. When the next round drops, previous themes have included AI for astronomical sciences, AI for materials research, and methods for strengthening AI itself. Future themes will likely follow the program's trajectory toward applied safety, autonomous discovery, and workforce development.

For general inquiries about the program, NSF directs researchers to AIInstitutesProgram@nsf.gov or (703) 292-5111. Program leads James Donlon and Erion Plaku in the CISE/IIS division manage the overall effort.

How to Position for the Next Round

Winning an AI Research Institutes award is not a solo effort. NSF expects multi-institutional, multi-disciplinary teams that create capabilities greater than the sum of their parts. If you are thinking about the next round, start building now:

Identify your theme early. Contact the relevant program officer to test whether your research direction fits the program scope. NSF recommends doing this at least five months before any submission deadline. The themes shift with each solicitation, so tracking the pattern -- from agriculture and weather to materials and AI safety -- helps you anticipate where the program is heading.

Build a genuine consortium. The funded institutes span three to ten partner institutions each. NSF reviewers look for meaningful integration, not a collection of loosely affiliated labs. Your proposal should demonstrate shared infrastructure, cross-institutional training programs, and research that requires the collaboration to work.

Plan for workforce development. Every institute includes education and workforce components. iSAT's 6,000-student reach and MMLI's training pipeline for molecular scientists are not afterthoughts -- they are central to the program's mission. Proposals that treat broader impacts as boilerplate lose ground to teams with concrete education plans.

Watch adjacent programs. The NAIRR Operations Center (NSF 25-546) is a related $35 million opportunity to build the national AI research infrastructure. The ExpandAI program (NSF 23-506) funds capacity building at institutions not yet part of the AI Institutes network. Both are worth tracking as stepping stones or complements to a full institute proposal.

The Takeaway

The AI Research Institutes program is NSF's flagship vehicle for large-scale, long-horizon AI investment -- and it shows no signs of slowing down. With 29 institutes now operating, a growing roster of private-sector co-funders, and strong alignment with federal AI priorities, the next solicitation could be the most competitive yet. The teams that start assembling their consortia and refining their research themes now will have a meaningful head start when NSF publishes the call.

If you are tracking NSF AI funding across multiple programs and deadlines, Granted can help you stay ahead of the solicitation cycle and build a stronger proposal from day one.