Drone sprays pesticides on wheat fields.
Artificial intelligence (AI) is on the cusp of fueling an agricultural revolution and helping to meet the challenge of sustainably feeding our growing world population. However, researchers warn that deploying new AI technologies at scale poses huge risks that are not being considered.
Imagine a field of wheat stretching to the horizon, grown for flour that is made into bread to feed the people of the cities. Imagine that all the authority to till, plant, fertilize, monitor and harvest this field has been delegated to artificial intelligence: algorithms that control drip irrigation systems, self-propelled tractors and combines intelligent enough to anticipate the weather and the environment respond to the exact needs of the culture. Then imagine a hacker messing things up.
“The idea of intelligent machines running farms is not science fiction.” — Asaf Tzachor
A new risk analysis recently published in the journal Nature Machine Intelligencewarns that the future use of artificial intelligence in agriculture poses significant potential risks to farms, farmers and food security that are poorly understood and underestimated.
“The idea of intelligent machines running farms is not science fiction. Big companies are already pioneering the next generation of autonomous Ag-Bots and decision support systems that will replace humans in the field,” said Dr. Asaf Tzachor of the University of Cambridge’s Center for the Study of Existential Risk (CSER), first author of the paper.
“But so far no one seems to have asked the question, ‘Are there any risks associated with rapid deployment of agricultural AI?'” he added.
Despite the great promise of AI to improve crop management and agricultural productivity, potential risks need to be responsibly addressed and new technologies need to be properly tested in experimental environments to ensure they are safe and protected from accidental failures, unintended consequences and cyberattacks, say the authors say.
In their research, the authors have developed a catalog of risks that must be taken into account in the responsible development of AI for agriculture – and ways to counter them. In it, they sound the alarm about cyber attackers who may be disrupting commercial AI farms by poisoning data sets or shutting down sprayers, autonomous drones and robotic harvesters. To prevent this, they suggest that “white hat hackers” help companies uncover security flaws during the development phase so systems can be protected from real hackers.
In a scenario involving accidental failure, the authors suggest that an AI system programmed only to produce the best crop yield in the short term could ignore the associated environmental consequences, leading to overuse of fertilizers in the long term and could lead to soil erosion. Overuse of pesticides in pursuit of high yields could poison ecosystems; Excessive application of nitrogen fertilizer would pollute the soil and surrounding waters. The authors suggest involving applied ecologists in the technology design process to ensure these scenarios are avoided.
Autonomous machines could improve the working conditions of farmers and relieve them of manual work. But without inclusive technology design, socioeconomic inequalities currently ingrained in global agriculture—including gender, class, and ethnic discrimination—will persist.
“Skilled AI farming systems that do not take into account the complexities of labor input will ignore and potentially perpetuate the exploitation of disadvantaged communities,” Tzachor warned.
Various Ag-Bots and advanced machines such as drones and sensors are already being used to collect information about crops and support farmers’ decision-making: for example to detect diseases or insufficient irrigation. And self-propelled combines can bring in a crop without the need for a human operator. Such automated systems aim to make farming more efficient, save labor costs, optimize production and minimize losses and waste. This leads to increased income for farmers as well as greater reliance on farming AI.
However, small farmers, who run the majority of farms worldwide and feed large parts of the so-called Global South, are likely to be left out of AI-related benefits. Marginalization, poor internet penetration rates, and the digital divide may prevent small farmers from taking advantage of advanced technologies, widening the gap between commercial and subsistence farmers.
With an estimated two billion people affected by food insecurity, including some 690 million malnourished and 340 million children with micronutrient deficiencies, artificial intelligence technologies and precision agriculture promise significant food security benefits in the face of climate change and a growing world population.
“AI is being hailed as a way to revolutionize farming. As we are using this technology extensively, we should carefully consider potential risks and aim to mitigate them early in the technology design,” said Dr. Seán Ó hÉigeartaigh, Executive Director of CSER and co-author of the new study.
Reference: “Responsible artificial intelligence in agriculture requires a systemic understanding of risks and externalities” by Asaf Tzachor, Medha Devare, Brian King, Shahar Avin and Seán Ó hÉigeartaigh, February 23, 2022, Nature Machine Intelligence.
DOI: 10.1038/s42256-022-00440-4
This research was funded by Templeton World Charity Foundation, Inc.
