The Potential Risks and Consequences of Implementing AI Agronomists for Smallholder Agriculture
Year: 2026
Type: article
Abstract: AI holds tremendous potential for improving the livelihoods of smallholder producers – whether through AI agronomy advice, improved crop genetics, better remote sensing for state incentive programs, or automation. AI won't solve all the problems for smallholder farmers (e.g. unclear land rights, distorting government incentives, lack of access to fertilizer), but it can be part of the larger solution. However, innovations in complex systems pose risks and can create unintended consequences. Mixing a powerful technology like a non-deterministic, non-explainable AI agronomy tool into a complex system responsible for the health and livelihoods of hundreds of millions of smallholder farmers deserves a concerted effort in anticipating, preventing, and mitigating the worst potential outcomes. This white paper makes a first attempt to catalog the potential negative consequences and harms of an AI agronomist on smallholder agriculture and offers a framework to identify potential mitigating actions. The framework can be applied to any type of AI-driven tool.
Author: Elliott, Grant
Institution: University of Cambridge
Language: English
Topic: Smart Agriculture and AI
Subfield: Plant Science
Domain: Life Sciences
Open Access status: green
OpenAlex ID: W7235690373
FWCI:
Cites:
Cited by:
Related to:
