top of page

The Rising Threat of AI: An Overview of the Material Influence of Technology on the Current Ecological Environment

by Sarya Nasser

Editor: Zosia Łukasiewicz

January 2025

Introduction

The growth and development of artificial intelligence (AI) are resulting in increased environmental strain because of the physical AI infrastructure and emissions that result from high data production. Companies are utilizing AI to progress their activities ranging through the fields of academia, medicine, entertainment, and e-commerce, among others (Peñarroya, 2024). Despite several advantages of AI, it is also accompanied by significant risks, such as environmental harm, discrimination and biases, cyber security vulnerabilities, and economic inequality (Thomas, 2024.). The deployment of AI technologies requires vast computational resources that are powered by energy-intensive data centers (Can We Mitigate AI’s Environmental Impacts?, 2024). These facilities consume large amounts of electricity while producing substantial harmful emissions. Additionally, the training of large AI models, such as natural language processors (computational devices that interpret and manipulate language (Stryker & Holdsworth, 2024))  and machine learning algorithms (self-programming computational algorithms), demands significant computational power, resulting in even higher emissions (Telefónica, 2023). This article will further comment on the current policies in place regarding AI regulation and will propose potential solutions to minimize ecological harm. The future of technology and AI depends on ensuring their use prudently and sustainably. 

 

Therefore, it is crucial to study ongoing perceptions and emerging environmental impacts of AI development while society is benefiting from the advantages of AI. This article will focus on the environmental impact associated with the increased use of AI, arguing that the common perception of AI as seamless and immaterial overlooks the ecological damage that the technology tends to generate.

Leveraging AI in Attempts to Tackle Environmental Issues: Advantages of AI 

 

Current uses of AI prove that the technology can revolutionize the way work is done across all industries (Peñarroya, 2024). Through the optimization of resource usage, AI is making processes more efficient at an exponential rate. With its autonomous and self-learning capabilities, AI has been a useful asset in advancing data modeling, facilitating green technologies, detecting crime, analyzing market data, optimizing learning, and forecasting future trends (Wang et al., 2024). The increased use of AI comes from its ability to optimize processes across various domains, saving time and resources. For instance, a study on the impact of AI in the agricultural industry in China shows that AI may serve as a significant tool in optimizing energy efficiency, lessening ecological footprints (Wang et al., 2024).  In the Netherlands, starting next month, the OWASIS (Observatory of Water Availability – System of Integrated Services) is projected to prevent flooding and droughts while improving water use management by using satellite data and AI-powered insights (Water Authorities Use Satellites to Monitor Drought, 2020). Moreover, the deployment of new infrastructures such as 5G, which make more sustainable use of resources than preceding versions ones, leads to a notable improvement in energy usage (Telefónica, 2023). These examples of AI-driven solutions convey how technology has the potential to optimize numerous challenges, including environmental issues. Across domains, AI has been a valuable asset in making processes faster and more cost-friendly; however, these advantages are not without a few drawbacks. 

 

Drawbacks of AI developments: Ecological Standpoint

Despite their seemingly immaterial nature, AI and cloud computing are built on large powerhouses and necessitate energy-intensive data centers for their maintenance. Most of the energy produced by these centers results from the burning of fossil fuels, a key contributor to global warming (Crownhart, 2024). AI systems are maintained through the extraction of data, labor, and earth’s minerals (Crawford, 2021). Technologies are intentionally designed not to last and become waste for financial gain, leading to a high turnover of material hardware. For instance, 75 percent of electronic devices are disposed of, rather than recycled (Devine, 2019). The supply chain of AI involves forms of exploitation of human labor and natural resources and massive concentrations of corporate and geopolitical power (Crawford, 2021). AI relies on many kinds of extraction: From harvesting the data made from users' digital interactions to depleting natural resources, AI depends on the exploitation of resources. With the use of AI, it is common to maximize computational cycles to improve performance (Crawford, 2021). This tendency toward “compute maximalism” has profound ecological impacts, increasing energy consumption and carbon emissions. Understanding the ecological footprint of AI is foundational for coming up with strategies to combat its risks and ensure its long-term sustainability.

 

Ecological AI Policies and Room for Improvement

To tackle the rising challenges as a result of AI, several regulations have been put in place. For instance, in attempts to make AI greener, the AI Environmental Impacts Act was developed, which outlines key provisions for mitigating environmental harm, including the formation of a consortium of stakeholders to evaluate the environmental impacts of AI  and a system for voluntary reporting by organizations on the ecological impacts of AI”  (Sen, 2024). The goals of the group of stakeholders, for instance, are to promote consistent and reliable reporting on the environmental impacts of AI, to encourage the use of open-source tools for measuring these effects, and to provide recommendations for reducing harm while promoting positive ones associated with AI technologies. This provision may prove to be the most impactful since it requires unified efforts. Ultimately, the Environmental Impacts Act aims to enhance the transparency, accountability, and understanding of the environmental footprint.

While some governance regarding AI is in place in terms of individual efforts, more comprehensive policies could play a key role in mitigating its environmental harm. While several tech giants such as Google Cloud Platform, Amazon Web Services, and Microsoft Azure have committed to achieving carbon neutrality by investing in sustainable energy projects, the lack of globally unified regulations poses a significant barrier. Through pledges to switch to renewable energy usage and reach carbon neutrality by certain timeframes, these providers are making innovations in AI while attempting to minimize ecological harm (Dua & Patel, 2024). However, energy consumption globally across data centers is projected to exceed 2000 TWh by 2035, challenging companies' aspirations of reaching their carbon neutrality targets (Moss, 2024). These findings highlight the difficulty of developing green AI in a rapidly expanding technological world. A key component for undergoing a significant environmental shift is a unified international effort to standardize green AI practices. A few challenges exist on the road towards this goal, including transnational incompatibilities regarding technological capabilities and political differences. Possible steps to achieving such a target may involve establishing a global certification program to certify and reward companies focused on sustainability and regulatory compliance or implementing continuous review programs whereby companies' adherence to global regulations is assessed. Ultimately, despite the potential of unified efforts to develop green AI, the rapid advancement of AI and global contrasts pose a few obstacles along the way.

Conclusion

While the advancement of AI provides new opportunities for ecologically sustainable development, challenges regarding their prudent use remain present, with several cases revealing its potential to do more harm than good. With a focus on the ecological influence associated with the increased use of AI, this article sought to provide an overview of the disadvantages of AI progress that are often masked by its advancements across multiple domains. This article also assessed the regulations in place regarding AI governance and discussed the advantages and disadvantages of a more unified global approach to minimize ecological harm. While the strive towards green AI development is projected to keep growing, efforts by regulators and industry players have been outpaced by the exponential rise and dynamism of the AI revolution. These observations highlight the urgency for unified regulatory action, which may be achieved through transnational reporting standards and rating systems, global certification programs, and continuous review cycles, while the exponential rise of AI is not seemingly slowing down. 

 

References

AI has an environmental problem. Here’s what the world can do about that. (2024, September 21). https://www.unep.org/news-and-stories/story/ai-has-environmental-problem-heres-what-world-can-do-about

Crawford, K. (2021). The Atlas of AI: Power, Politics, and the Planetary Costs of Artificial Intelligence. Yale University Press.

Data, Digital Technology, and the Environment. (2024, December 3). https://www.genevaenvironmentnetwork.org/resources/updates/data-digital-technology-and-the-environment/

Devine, K. (2019). Decomposed: The Political Ecology of Music. The MIT Press. https://doi.org/10.7551/mitpress/10692.001.0001

Telefónica. (2023, September 13). How does technology affect the environment? https://www.telefonica.com/en/communication-room/blog/how-technology-affect-environment/

 

Moss, S. (2024, October 16). IEA: Global data center electricity consumption to “increase significantly,” but remain a small part of overall usage. Data Center Dynamics. https://www.datacenterdynamics.com/en/news/iea-global-data-center-electricity-consumption-to-increase-significantly-but-remain-a-small-part-of-overall-usage/

 

Rillig, M. C., Mansour, I., Hempel, S., Bi, M., König-Ries, B., & Kasirzadeh, A. (2024). How widespread use of generative AI for images and video can affect the environment and the science of ecology. Ecology Letters, 27(3), e14397. https://doi.org/10.1111/ele.14397

 

Roundy, J. (2023, July 12). Assess the environmental impact of data centers | TechTarget. Search Data Center. https://www.techtarget.com/searchdatacenter/feature/Assess-the-environmental-impact-of-data-centers

Sen, H. Y. (2024, March 13). The Artificial Intelligence Environmental Impacts Act of 2024: What You Need to Know. Holistic AI. https://www.holisticai.com/blog/artificial-intelligence-environmental-impacts-act

Stryker, C., & Holdsworth, J. (2024, August 11). What is NLP (natural language processing)? IBM. https://www.ibm.com/think/topics/natural-language-processing

The EU AI Act: Insights from the Green AI Committee. (2024, October 17). Green Software Foundation. https://greensoftware.foundation/articles/the-eu-ai-act-insights-from-the-green-ai-committee

Wang, Y., Zhang, R., Yao, K., & Ma, X. (2024). Does artificial intelligence affect the ecological footprint? –Evidence from 30 provinces in China. Journal of Environmental Management, 370, 122458. https://doi.org/10.1016/j.jenvman.2024.122458

Water authorities use satellites to monitor drought. (2020, June 23). Netherlands Space Office. https://www.spaceoffice.nl/en/news/378/water-authorities-use-satellites-to-monitor-drought.html

bottom of page