Burlington, Vt. — Feb 25, 2026 — A new peer-reviewed study co-authored by Virginia Tech and University of Vermont researchers offers one of the first, large-scale empirical looks at how Certified Crop Advisors (CCAs) across North America evaluate the next generation of artificial intelligence–enabled decision support systems (AI‑DSS) for agriculture. Published in Technological Forecasting and Social Change (Elsevier), the study identifies the specific design features that most influence whether trusted agricultural advisors will choose AI tools—and what might hold them back.
The research that was conducted in collaboration with the American Society of Agronomy was led by Maaz Gardezi an Associate Professor in the School of Public and International Affairs at Virginia Tech, with co-authors from UVM: Professor Asim Zia, Professor Donna M. Rizzo, Research Associate Professor Scott C. Merril, UVM graduate students Benjamin E.K. Ryan and Halimeh Abuayyash, and Virgina Tech Graduate Students Indunil Dharmasiri, Pablo Carcamo, Bhavna Joshi. Additional collaborators were David Clay Distinguished Professor at South Dakota State University and John McMaine Extension Associate Professor at University of Kentucky. The research team used a discrete-choice experiment to analyze how crop advisors weigh trade-offs among cost, accuracy, spatial precision, and data ownership when evaluating AI-based systems.
Key Findings
- Simplicity and usability matter most. Advisors consistently favored systems that were easy to use and incorporated satellite data over more complex, ultra-accurate tools requiring intensive data inputs.
- Trust depends on transparency and data governance. Cost and data ownership emerged as major determinants of adoption, with advisors preferring systems that allow users to retain full or shared control over their data.
- AI shouldn't replace professional judgment. Crop advisors favored AI-DSS tools that augment rather than automate their work, valuing editable recommendations, local calibration, and field-verification options.
- Tech attitudes shape adoption. Advisors with more optimistic views of AI were more open to data-intensive systems, while those with privacy concerns were less likely to adopt tools that require extensive farmer data.
Maaz Gardezi the study's PI summarizes the important insight of the research in this way, "Technical performance of AI tools matters in agriculture, but cost and data ownership—especially shared or open models—are pivotal to selection. Crop advisors prefer systems that augment rather than replace professional judgment."
A Turning Point for Agricultural AI
The study arrives at a time when AI-generated predictions, classifications, and recommendations are increasingly used to guide decisions involving fertilizer application, pest and disease management, irrigation scheduling, and carbon and nutrient accounting. Yet adoption has lagged, especially among mid-sized and smaller farms, due partly to concerns about privacy, affordability, transparency, and trust.
"Certified crop advisors are among the most trusted technical experts that farmers in the US turn to," said Asim Zia, Professor of Public Policy and Computer Science at UVM. "Designing AI decision tools that enhance, not replace, their expertise is essential for building agricultural systems that are productive, equitable, and climate‑resilient."
A Socio-Technical Framework for Trustworthy AI
The authors argue that AI developers and policymakers must take a socio-technical approach that aligns algorithms with the real-world values and constraints of the people expected to use these tools. Their findings recommend:
- Co-creation with crop advisors and farmers during development
- Transparent cost structures and clear communication of trade-offs
- User-controlled data governance models
- Human-in-the-loop designs that preserve advisor autonomy
"These insights help move AI for agriculture beyond performance metrics," said study co-author Donna Rizzo, Dorothean Chair and Professor of Civil & Environmental Engineering at UVM. "The goal is trustworthy, context-sensitive tools that work for diverse farms and advisory systems."
About the Study
The article, " A socio-technical framework for analyzing crop advisors' preferences for AI-based decision support systems ," appears in the May 2026 issue of Technological Forecasting and Social Change. The research was supported by the National Science Foundation (Grant Nos. 2202706 and 2026431) and the USDA National Institute of Food and Agriculture (Award No. 2023‑67023‑40216).
