Content
April 29, 2026
Generative artificial intelligence (AI) is increasingly recognized as a tool for improving farm decision-making. Farming requires continuous decisions about planting, crop management, including fertilizer use, chemical use, irrigation, and pest control, all of which are made under uncertainty from weather, markets, and biological risks. Current decision-support systems often address one issue at a time, can be difficult to use, time intensive, and limited in their ability to integrate multiple data sources. This study investigates whether generative AI, specifically ChatGPT – 4o, can serve as a more comprehensive, accurate and adaptive decision-support tool in real-world crop production.
Using the University of Nebraska–Lincoln’s TAPS farm management competition, a University of Nebraska-Lincoln research group tested AI-assisted decision-making alongside experienced producers managing similar plots. The AI system was provided with farm-level data, including soil conditions, weather information, and historical performance, and was used to generate decisions across the full crop production cycle. Results show that AI-generated recommendations were logical, timely, and operationally feasible. The AI managed plot achieved above-average yields, ranking in the top third of all participants and producing statistically higher yields than the average farmer-managed plots.
Despite strong agronomic performance, the study identifies important tradeoffs that have direct policy relevance. While AI improved yield outcomes, it was not the best performer in input efficiency or profitability. Irrigation water-use efficiency was lower than average, and economic returns were constrained by relatively simple grain marketing decisions, which were not decided by AI. These findings suggest that while AI can effectively support production decisions, it must be integrated with economic optimization tools and market information systems to deliver full value to producers.
An important point of the study is that data availability and quality are critical to AI performance. The system performed well when provided with structured, relevant data, but struggled when key inputs, such as real-time weather forecasts, irrigation costs, or spatial field variability, were missing. The lack of standardized, accessible, and interoperable agricultural data systems indicates an important policy challenge. Investments in rural data infrastructure, including sensor networks, data platforms, and broadband connectivity, will be essential to enable effective deployment of AI decision-support tools.
The study also underscores the continued importance of human oversight. AI recommendations were occasionally adjusted by researchers based on contextual knowledge or emerging conditions, demonstrating that expert opinion remains necessary. This has implications for extension services and workforce development. Rather than replacing human expertise, AI is more likely to augment decision-making, increasing the demand for digitally skilled producers, agronomists, and advisors who can interpret and refine AI outputs.
Adoption barriers exist. Prior research and findings from this study indicate that farmers may be reluctant to rely on algorithm-based tools due to concerns about trust, transparency, cost, and usability. Many producers continue to prefer advice from human subject matter experts, even when automated tools provide accurate recommendations. Addressing these concerns will require policies that promote transparency in AI systems and demonstrate clear economic benefits to producers.
Generative AI has the potential to transform agricultural decision-making by integrating multiple data streams and generating increasingly sophisticated recommendations. Future systems may produce field-level management maps, automate irrigation and fertilizer applications, and incorporate real-time market and weather data into fully integrated decision frameworks. Realizing this potential will require coordinated investments in research, data systems, and extension programs, as well as careful attention to equity and accessibility to ensure that resource-constrained producers are not left behind.
Generative AI represents a significant opportunity to enhance productivity and decision-making in agriculture, but it is not yet a standalone solution. Policymakers should view AI as a complementary tool that, when combined with human expertise and robust data infrastructure, can improve farm performance and resource use. Strategic investments in data systems, extension, and applied research will be critical to unlocking the full benefits of AI for agricultural producers and rural communities.
Link to the study
https://www.sciencedirect.com/science/article/pii/S2589721726000255
Journal publication
Chamara, N., Pan, Y., Taghvaeian, S., Walters, C., Proctor, C., Rudnick, D., Redfearn, D., Luck, J. and Ge, Y., 2026. Can generative AI make farming decisions? Current status and future pathways–A case study in row crop production with ChatGPT. Artificial Intelligence in Agriculture.
NotebookLM podcast
https://notebooklm.google.com/notebook/02e4ecd5-ee39-43d1-87cb-9c8c85fb323d/artifact/34370527-900b-4223-8392-8c7448d53284
Cory Walters
Associate Professor
Department of Agricultural Economics
University of Nebraska-Lincoln
cwalters7@unl.edu