Grants for AI Projects: Where to Find AI Funding in 2026

Grants for AI projects fall into three buckets: federal agency programs, state and local innovation funds, and private or nonprofit AI-for-good grants. Each bucket has different rules, award sizes, and who can win.

This guide covers real, verified programs that exist in 2026. It scopes to United States funding only. It explains how AI funding is structured, names specific programs by government level, and shows the eligibility trap that stops most applicants before they start.

One honest caveat up front: most federal grants for AI projects require a research mission or a qualifying small-business structure. A pure services company often does not qualify. We say plainly below who does.

This guide pairs with a downloadable dataset of AI grant programs — award sizes, agencies, and links — so you can filter by fit. The download is handled separately on this page.

How AI Grant Funding Is Structured

AI grant funding in the United States is structured across three levels: federal, state and local, and private or NGO. Each level funds a different kind of work and a different kind of applicant.

Federal money is the largest pool. It flows through science agencies for research, and through the SBIR/STTR programs for small businesses turning research into products. Most federal opportunities are posted on Grants.gov.

State and local money is smaller but easier to reach. It usually rewards local job creation, university partnerships, or matching a federal award you already won.

Private and NGO money is the most flexible. Foundations and big-tech AI-for-good programs fund nonprofits and mission-driven work, often with cloud credits and mentorship on top of cash.

• Federal — research grants and small-business commercialization funding; largest awards, most red tape.
• State and local — economic-development and innovation grants; smaller, local-benefit focused.
• Private and NGO — foundation and corporate AI-for-good grants; flexible, often nonprofit-only.
• Discovery layer — federal opportunities are searchable on Grants.gov; private ones on each funder's site.

Federal Grants for AI Projects

Federal grants for AI projects come mainly from science agencies and the small-business seed programs. The two biggest doors are the National Science Foundation and the SBIR/STTR programs it runs.

The NSF National AI Research Institutes program funds large, multi-year university-led research centers, typically around $20 million each over four to five years. This is research funding, not startup funding — universities and research organizations lead it.

For companies, the real door is SBIR/STTR, branded as America's Seed Fund. Its AI topic awards up to about $305,000 in Phase I and up to $1.25 million in Phase II. This money is non-dilutive — you do not give up equity.

Other federal agencies fund AI too. The National Institutes of Health funds AI in health and biomedicine, the Department of Energy funds AI for science and the grid, and the National Endowment for the Humanities funds humanities-and-AI research, including centers up to $750,000.

Here is the eligibility trap almost no one explains. SBIR/STTR requires a U.S. small business (under 500 employees) with at least 51% ownership by U.S. citizens or permanent residents, and it will not fund firms majority-owned by multiple VC, private-equity, or hedge funds. It funds technology development, not client services — so an AI consultancy billing hourly usually does not qualify on services work alone.

• NSF National AI Research Institutes — university-led research centers, ~$20M each; for research organizations, not startups.
• NSF SBIR/STTR (America's Seed Fund), AI topic — up to ~$305K Phase I, up to $1.25M Phase II; non-dilutive; U.S. small businesses only.
• NIH — AI in health, diagnostics, and biomedical research; for research teams and eligible small businesses.
• DOE — AI for scientific computing, energy, and the grid.
• NEH — humanities perspectives on AI, including research centers up to $750K.
• Where to find them — search and apply through Grants.gov; SBIR opportunities are also listed on SBIR.gov.

Federal AI Program Comparison Table

Use this table to match a federal program to who you are and what you are building. Each row lists the program, who it is for, and a typical award.

Program — Who it is for — Typical award:

• NSF National AI Research Institutes — universities and research organizations — about $20M over 4 to 5 years.
• NSF SBIR/STTR AI topic (America's Seed Fund) — U.S. small businesses building AI technology — up to ~$305K (Phase I), up to $1.25M (Phase II).
• NIH AI programs — biomedical researchers and eligible small businesses — varies by institute and mechanism.
• DOE AI programs — labs, universities, and energy-focused firms — varies by office and solicitation.
• NEH Humanities Research Centers on AI — universities and independent research organizations — up to $750K per center.

State & Local AI Funding

State and local AI funding usually comes from economic-development agencies and university tech-transfer offices. These grants reward local job creation, commercialization, and partnerships with in-state universities.

A common and useful category is the state SBIR match. Several states top up a federal SBIR/STTR award with their own cash. For example, Empire State Development in New York runs NYSTAR Innovation Matching Grants that add up to $100,000 (Phase I) or $200,000 (Phase II) to businesses that already won federal SBIR/STTR funding.

University tech-transfer and innovation funds are the other main category. Many public universities run proof-of-concept or gap funds that back AI research on a path to a spinout or license. These vary widely by school, so check the tech-transfer office where your research lives.

The practical move is to search your own state's economic-development site for words like innovation, matching, technology, or commercialization grant. Programs differ by state, so treat the New York example as a template for the category, not a nationwide offer.

• State SBIR match — extra state cash for firms that already won a federal SBIR/STTR award (e.g. NYSTAR Innovation Matching Grants).
• State innovation and tech grants — economic-development funding tied to local jobs and commercialization.
• University tech-transfer funds — proof-of-concept and gap funding for campus AI research heading to market.
• How to find them — search your state economic-development agency and your local university tech-transfer office.

Private & NGO AI Grants

Private and NGO AI grants come from foundations and big-tech AI-for-good programs. They are the most flexible source and often the best fit for nonprofits and mission-driven teams.

Google.org runs AI-focused impact challenges and a generative-AI accelerator that give selected nonprofits cash, Google Cloud credits, and pro-bono engineering help. Award sizes reach into the millions on its larger challenges.

The Patrick J. McGovern Foundation is one of the largest funders of public-purpose AI, granting tens of millions each year to nonprofits and research groups. Note it does not take unsolicited applications, so most funding starts as a relationship, not a cold form.

Big-tech programs round out the field. AWS Imagine Grant funds U.S. 501(c) nonprofits with cash, AWS credits, and technical support, including a track for frontier-AI workloads. Microsoft AI for Good runs open calls that pair grants and Azure credits with its research lab. Fast Forward accelerates tech nonprofits — many now AI-powered — with unrestricted funding and mentorship. Mission-driven funders like the Mozilla Foundation also back trustworthy-AI work.

• Google.org — AI impact challenges and a generative-AI accelerator; cash plus Google Cloud credits and engineering help.
• Patrick J. McGovern Foundation — large public-purpose AI grants; relationship-based, no unsolicited applications.
• AWS Imagine Grant — U.S. 501(c) nonprofits; cash, AWS credits, and a frontier-AI track.
• Microsoft AI for Good — open-call grants plus Azure credits and research-lab collaboration.
• Fast Forward — accelerator for tech and AI nonprofits; unrestricted funding and mentorship.
• Mozilla Foundation — funding and programs for trustworthy, open, public-interest AI.

How to Apply for AI Grants

You apply for most federal AI grants through Grants.gov, but you cannot submit until your organization is registered first. Registration takes real time, so start it before you find a deadline you like.

Federal registration runs through SAM.gov, which issues your Unique Entity Identifier (UEI). Budget two to four weeks for SAM.gov to go active, and start at least three weeks before any submission deadline. Registration is free and must be renewed each year.

Eligibility basics decide everything. Confirm your entity type — small business, nonprofit, or research organization — matches the program before you write a word. For SBIR/STTR, confirm the U.S.-ownership and small-business rules apply to you.

Plan the full timeline realistically. From spotting an opportunity to usable money, federal grants commonly take four to nine months. Private grants move faster but run on fixed cohort or challenge windows.

• Step 1 — Register on SAM.gov and get your UEI; allow 2 to 4 weeks.
• Step 2 — Create a Grants.gov applicant account and link it to your organization.
• Step 3 — Confirm eligibility (entity type, ownership, mission) before writing.
• Step 4 — Match one program per proposal; do not force-fit a mismatched one.
• Step 5 — Plan 4 to 9 months for federal timelines; watch fixed windows for private grants.

AI Grant Funding Levels Compared

Compare the three funding levels on award size, who qualifies, effort, and speed before you commit. Each level suits a different applicant.

Level — Typical award — Best for — Effort and speed:

• Federal (SBIR/STTR) — ~$305K to $1.25M+ — U.S. small businesses building AI tech — high effort, 4 to 9 months, non-dilutive.
• Federal (research grants) — up to ~$20M — universities and research organizations — high effort, long timelines.
• State and local — ~$25K to $200K — in-state firms, SBIR winners, campus research — medium effort, faster than federal.
• Private and NGO — ~$25K to several million (often plus credits) — nonprofits and mission-driven teams — medium effort, fixed windows.

How to Strengthen an AI Grant Application

You strengthen an AI grant application by matching the funder's exact mission and proving real-world impact, not just technical novelty. Reviewers fund outcomes, not model architectures.

Lead with the problem and who it helps, then show the AI as the means. State clear, measurable results the money will produce inside the grant term.

Be specific about data, ethics, and risk. Programs increasingly want to see how you handle privacy, bias, and safety — a vague answer here sinks otherwise strong proposals.

For a step-by-step walkthrough of drafting, structuring, and using AI to write the narrative faster, see our companion guide on AI grant writing. It covers the parts of the application that most often decide the outcome.

• Match the mission — mirror the funder's stated priorities in your first paragraph.
• Show impact — name measurable outcomes, not just capabilities.
• Address risk — cover data, privacy, bias, and safety directly.
• Fit the budget — request an amount inside the program's stated range.
• Go deeper — see /guides/ai-grant-writing for the full drafting playbook.