USDA's Quiet AI Revolution: Machine Learning Grants for Agriculture

Most researchers hunting for AI funding think NSF or DARPA. They should be looking harder at USDA.

The department has quietly assembled one of the federal government's most expansive portfolios of applied machine learning investment — spread across multiple agencies, spanning crop science to food safety — and much of it remains undersubscribed relative to the sophistication of the problems it addresses. For university research teams, ag-tech startups, and land-grant institutions, the opportunity is significant and the competition is less fierce than in the places everyone else is looking.

The NIFA Backbone: AFRI's Data Science Program

The primary grant vehicle is USDA's National Institute of Food and Agriculture (NIFA), through the Agriculture and Food Research Initiative's Data Science for Food and Agricultural Systems program (A1541, known internally as DSFAS). NIFA has invested roughly $7–$7.6 million annually in A1541 in recent cycles, with individual grants capped at $650,000 for standard awards over project periods up to five years.

The scope is deliberately broad. Projects must be "equally well-grounded in agricultural sciences and in data science or AI," which means teams need real depth on both sides — agricultural domain expertise and methodological rigor. That requirement is a filter, not a barrier: it's also what makes the program genuinely interesting. Funded work has examined how ML models can optimize farm-level decisions, how AI can be applied to agricultural supply chains, and how data science tools can help manage natural resources.

For larger collaborative efforts, NIFA offers Coordinated Innovation Networks (CIN) at up to $1,000,000 per project — and a specific CIN-FM track dedicated entirely to food supply chain modeling. Applications in the CIN-FM track focus on transitions to "robust, resilient, and cooperative food supply networks," an area that has drawn renewed attention after the supply disruptions of the early 2020s.

The FY26 AFRI Foundational and Applied Science RFA has closed its A1541 window (deadline was December 18, 2025), but NIFA releases new RFAs each year. Researchers should monitor NIFA's funding opportunities page and subscribe to notifications — the FY27 cycle will open in late summer 2026. Contact: nifa-dsfas@usda.gov.

Five National Institutes, One Interconnected Research Network

A level above individual grants, NIFA co-funds — alongside NSF — a network of five National AI Research Institutes dedicated to agriculture, part of a broader $220 million federal investment in AI research institutes.

The five institutes cover the field aggressively:

• AgAID (Washington State University) — AI for farm workforce and decision support, autonomous systems, and robotics in specialty crops
• AIFARMS (University of Illinois Urbana-Champaign) — foundational AI for autonomous farming, livestock management, and soil health
• AIFS (University of California, Davis) — next-generation food systems and sustainable supply chains
• AIIRA — focused on agricultural AI for climate resilience
• AI-LEAF — AI applications for sustainable forestry and ecosystems

These institutes are not grant opportunities in themselves, but they define the research frontier that NIFA's competitive grants respond to. Teams affiliated with or adjacent to an institute often have an easier path to articulating their work's significance — and the institutes actively collaborate with external researchers on specific projects.

ARS and the AI-COE: Agency-Internal Innovation Funds

The USDA Agricultural Research Service runs a parallel track through its AI Center of Excellence (AI-COE), housed within the USDA Scientific Computing Initiative (SCINet). The AI-COE funds internal ARS scientists directly — 4 to 6 awards per cycle at up to $100,000 each.

The program paused in FY25 and returned for FY26 with a February 6, 2026 deadline. Eligible applicants are ARS Category 1, 4, or 6 scientists. Projects must develop or adapt an AI/ML method to answer a specific scientific question, or create a prototype digital tool for producers or researchers. Working groups and training activities are explicitly excluded — the program wants working prototypes and publishable methods.

In practice, AI-COE projects have addressed cover crop and weed detection using computer vision, crop phenotyping automation via UAV imagery, and breeding decision support using high-performance computing. The Breeding AI and ML Working Group within SCINet connects researchers across locations working on similar problems.

For institutions with existing ARS partnerships or Cooperative Research and Development Agreements, the AI-COE is a direct path into federally supported applied ML work with production agriculture applications.

What USDA Actually Wants to Fund

The breadth of the portfolio is sometimes obscured by the department's bureaucratic structure. Across NIFA and ARS, the research agenda converges on several applied areas:

Precision agriculture and autonomous systems. USDA-funded projects have explored machine learning for soil health monitoring, remote sensing for crop stress detection, and autonomous robots for labor-intensive tasks like harvesting and targeted herbicide application. One widely cited example from ARS-funded work: weed detection algorithms that combine computer vision with robotic sprayers to dramatically reduce herbicide use.

Food safety. NIFA's food safety priority area supports AI applied to pathogen detection, contamination source tracing, and smart packaging systems that monitor product condition through the cold chain. An earlier DSFAS cycle funded a University of Illinois project on big data analytics for microbiological risk in fresh produce, and a Texas A&M project on disease and pest management integrated with agricultural logistics.

Supply chain and market modeling. The CIN-FM track, and the AIFS institute more broadly, focus on computational modeling of food distribution networks — demand forecasting, transportation optimization, and resilience stress-testing. This is the area where agricultural economics and machine learning overlap most directly.

Crop modeling and climate resilience. NIFA supports AI-enhanced crop models that incorporate weather variability, soil data, and phenological observations to predict yield outcomes. As Secretary Rollins' December 2025 research priorities memorandum makes clear, farm profitability — including through mechanization and precision inputs — is a top-tier departmental priority.

Building a Proposal That Lands

A common mistake is treating USDA AI grants as pure computer science funding with an agricultural wrapper. NIFA evaluates proposals on whether the team has genuine agricultural domain knowledge, not just technical capacity. Co-investigators from land-grant extension programs, USDA research stations, or commodity organizations carry real weight.

The DSFAS program explicitly requires both grounding in agricultural science and grounding in data science or AI. Teams that lead with the ML methodology and add agriculture as an application domain often score lower than teams that lead with a specific agricultural problem — reduced input costs, improved food safety detection, more accurate yield forecasting — and then demonstrate that machine learning is the right tool to solve it.

Coordinated Innovation Network applications also require sustainability plans: how will this research program continue beyond the grant period? Industry partnerships, extension relationships, and commercialization pathways all strengthen the case.

For researchers navigating USDA's distributed funding landscape — across NIFA program areas, ARS innovation funds, and the AI institutes ecosystem — Granted gives you a filtered view of what's open, what's closing, and where your work fits.

Sources:

• Data Science for Food and Agricultural Systems (DSFAS) | NIFA
• Artificial Intelligence | NIFA
• ARS AI Innovation Fund (FY26) | SCINet
• USDA-NIFA and NSF Invest $220M in Artificial Intelligence Research Institutes | NIFA
• NIFA Invests $7.4M in Data Science for Food and Agricultural Systems | NIFA
• AFRI Deadlines | NIFA
• Secretary Rollins Announces New Priorities for Research and Development in 2026 | USDA
• United States Department of Agriculture Fiscal Year 2025-2026 AI Strategy | USDA
• Advancing AI in Agriculture through Large-Scale Collaborative Research | Communications of the ACM