The Human and Environmental Cost of AI

Authored by: Ryan Nisay

Art by: Mia Hsu

Artificial Intelligence (AI) is already being hailed as humanity's next greatest technological leap, but behind its promise lies a growing public health crisis. In less than a decade, what began as an abstract concept confined to research laboratories has essentially embedded itself in every sphere of human activity. However, the massive computational power fueling AI systems demands staggering amounts of electricity, most of which still comes from fossil fuels. Furthermore, AI systems also impact every aspect of public health, and in ways that are slowly fueling a new public health crisis. The world's increasing dependence on AI has quietly created an ecosystem of energy-intensive data centers, ballooning electricity demands, and, as recent research reveals, escalating threats to public health. 

AI, by itself, is not the root cause of the public health crisis; rather, the complications surrounding AI's development and use significantly impact human health. Adam Wierman and Shaolei Ren, professors at the California Institute of Technology and the University of California, Riverside, respectively, highlight the public health crisis of AI by linking it to the environmental crisis [1]. Training large language models requires a considerable amount of electricity, most of which is generated from fossil fuels. Increased energy use releases excess fine particulate matter and nitrogen oxides, which are associated with an increased risk of respiratory and cardiovascular diseases [2]. By 2030, Wireman and Ren estimate that emissions from AI data centers could cause over 1,300 premature deaths in the United States each year, with the total public health costs reaching upwards of $20 billion [3]. 

The public health crisis extends beyond air pollution and energy consumption. AI infrastructure also places extreme stress on local water infrastructure. In Querétaro, Mexico, newly constructed data centers consume approximately 15 billion liters of water annually, accounting for about 13% of the metropolitan area's total water use [4]. This overconsumption by data centers creates competition for water among local communities where they are built, harming agriculture, industry, and the community's overall health. Agriculture around Querétaro relies on irrigation from the same aquifers that are now stressed by industrial and data center withdrawals. 

Furthermore, the entire lifecycle of AI causes a significant environmental burden. Extracting and processing rare earth metals for AI hardware not only generates thousands of tons

of toxic waste per ton of ore but also involves labor practices that expose workers to hazardous chemicals, environmental contamination, and human rights violations [5]. For instance, lithium exposure—common in mining regions—can contaminate water sources and lead to neurological, kidney, and reproductive damage in humans and animals [6]. AI's promise of progress is built on exploitative systems that harm the most vulnerable populations. The environmental and social costs embedded within each chip, sensor, and data center ultimately put public health at risk. 

The public health implications of AI extend beyond the physical environment. Pollution,

ecosystem degradation, and climate change have been increasingly linked to rising rates of anxiety, depression, and psychological distress [7]. The mental health crisis can be exacerbated by environmental decline brought on by AI's environmental footprint. AI's environmental footprint may further exacerbate chronic diseases already tied to pollution and climate change. The effects of AI are not confined solely to the physical or environmental aspects; they permeate all aspects of human life. 

At the center of 21st-century innovation, AI’s rapid integration into society offers profound potential to advance all aspects of life; however, these gains come at a cost we can no longer ignore. The environmental destruction, human rights violations, and public/mental health consequences linked to AI development expose a dangerous rift between innovation and sustainability. Actual progress demands sustainable integration—not only for our environment but also for the health of our world. If AI is humanity’s next greatest technological leap, it must also become its most ethically guided one; an advancement made not at the expense of the world that sustains it.

References: 

1. Wierman, A. (2024). Air Pollution and the Public Health Costs of AI. California Institute of Technology. 

https://www.caltech.edu/about/news/air-pollution-and-the-public-health-costs-of-ai 

2. Du, Y., Xu, X., Chu, M., Guo, Y., & Wang, J. (2016). Air particulate matter and cardiovascular disease: the epidemiological, biomedical and clinical evidence. Journal of thoracic disease, 8(1), E8–E19. https://doi.org/10.3978/j.issn.2072-1439.2015.11.37 

3. Han, Y., Wu, Z., Li, P., Wierman, A., & Ren, S. (2024, December 9). The Unpaid Toll: Quantifying the Public Health Impact of AI. ArXiv. https://arxiv.org/abs/2412.06288 

4. 4. Fiske, A., Radhuber, I. M., Willem, T., Buyx, A., Celi, L. A., & McLennan, S. (2025). Climate change and health: the next challenge of ethical AI. The Lancet Global Health, 13(7), e1314–e1320. https://doi.org/10.1016/S2214-109X(25)00124-X 

5. 5. Ligozat, A.L., Lefevre, J., Bugeau, A., & Combaz, J. (2022). Unraveling the Hidden Environmental Impacts of AI Solutions for Environment Life Cycle Assessment of AI Solutions. Sustainability, 14(9). https://doi.org/10.3390/su14095172 

6. Jaishankar, M., Tseten, T., Anbalagan, N., Mathew, B. B., & Beeregowda, K. N. (2014). Toxicity, mechanism, and health effects of some heavy metals. Interdisciplinary toxicology, 7(2), 60–72. https://doi.org/10.2478/intox-2014-0009 

7. Baker, E., Cynthia Faye Barlow, Daniel, L., Morey, C., Bentley, R., & Mark Patrick Taylor. (2024). Mental health impacts of environmental exposures: A scoping review of evaluative instruments. Science of the Total Environment, 912. 

https://doi.org/10.1016/j.scitotenv.2023.169063