The hidden side of digital transformation in agriculture: Artificial intelligence and carbon emissions

A research team including Assoc. Prof. Mesut Öztırak, a faculty member at Istanbul Medipol University School of Business and Management Sciences, examined the environmental impact of artificial intelligence use in agriculture. Published in the journal Sustainable Development, the study entitled “Artificial Intelligence in Agriculture: A Pathway to Improved Environmental Outcomes?” analyzed data from the nine countries that invested most heavily in artificial intelligence between 2012 and 2023.

The findings revealed that as investment in artificial intelligence in agriculture increased, carbon emissions could also rise, with this relationship being more pronounced in countries with higher emission levels.

ARTIFICIAL INTELLIGENCE CAN MAKE AGRICULTURE MORE EFFICIENT
Artificial intelligence can be used in numerous areas of agriculture, including irrigation, fertilization, pesticide application, production planning, and environmental monitoring. Sensors and smart systems can help determine more accurately the amount of water or fertilizer required, thereby enabling more efficient use of resources.

However, the study shows that the environmental impact of technology cannot be explained solely by the efficiency gains it provides. Artificial intelligence applications focused particularly on increasing production may lead to greater consumption of energy and production inputs. The energy consumption required to operate the technological infrastructure underlying artificial intelligence systems is also considered part of their environmental impact.

TWELVE YEARS OF DATA FROM NINE COUNTRIES WERE ANALYZED
The study examined Canada, China, Estonia, France, India, Israel, Japan, Switzerland, and the United States. Data from these countries between 2012 and 2023 were compared to investigate the relationship between artificial intelligence investment and agricultural carbon emissions.

In addition to artificial intelligence investment, the study considered various factors, including economic growth, rural population, and countries’ openness to international trade. The researchers employed two different statistical analysis methods to identify variations across countries and across different emission levels.

A NOTABLE RELATIONSHIP BETWEEN ARTIFICIAL INTELLIGENCE INVESTMENT AND EMISSIONS
The study’s most notable finding was that increased investment in artificial intelligence was accompanied by higher agricultural carbon emissions. The researchers emphasize that this finding does not mean that artificial intelligence is inherently harmful to the environment. Rather, the results underscore the importance of using technology not only to increase production but also to reduce resource consumption and environmental impacts.

Within this context, applications such as smart irrigation, precision agriculture, and data-driven fertilizer and pesticide use stand out as areas in which artificial intelligence could deliver environmental benefits.

THE RELATIONSHIP IS STRONGER IN HIGH-EMISSION COUNTRIES
Another important finding of the study was that the relationship between artificial intelligence investment and carbon emissions varied according to countries’ existing emission levels. In countries with high emission levels, the relationship between artificial intelligence investment and emissions was stronger.

This finding suggests that applying the same artificial intelligence policies to every country’s agricultural sector may not be sufficient. A technology that delivers environmental benefits in one country may not produce the same outcome in another with different production conditions. Artificial intelligence investments should therefore be planned with countries’ economic structures, agricultural production practices, and environmental policies taken into consideration.

EMISSIONS ALSO RISE AS THE RURAL POPULATION INCREASES
The study found that agricultural carbon emissions were also higher in countries with larger proportions of rural populations. The researchers associate this finding with the prevalence of traditional and less efficient agricultural practices. It is therefore important to support education, infrastructure, and environmentally friendly practices alongside technological investment in rural areas.

By contrast, economic growth and openness to international trade were found to be associated with lower agricultural carbon emissions. The findings suggest that economic development may facilitate the adoption of more environmentally friendly production technologies, while international trade may contribute to the wider adoption of environmental standards and more efficient production practices.

ENVIRONMENTAL GOALS SHOULD BE PRIORITIZED IN ARTIFICIAL INTELLIGENCE INVESTMENT
The study demonstrates that investment in artificial intelligence in agriculture should not be evaluated solely in terms of production and efficiency. It recommends that technological investments be planned in conjunction with environmental sustainability goals, that training and advisory opportunities be expanded to enable farmers to adopt new technologies, and that the necessary data infrastructure be strengthened.

THE STUDY’S CONTRIBUTION TO THE LITERATURE
The study is significant in that, rather than examining the general relationship between artificial intelligence and the environment, it focuses specifically on the relationship between artificial intelligence investment in the agricultural sector and agricultural carbon emissions. By also demonstrating how this relationship varies across different emission levels, the study shows that assessments of digital transformation in agriculture should consider not only increases in production and efficiency but also the environmental consequences of technology.