Introduction

The demand for AI has skyrocketed in recent years, with many industries advocating for its expansion into their workflows. As this demand grows, so does the power consumption of the data centers, which are the backbone of AI. This requires an expanding electrical power infrastructure, which is affecting the communities around it. This demand is also affecting the global sustainable energy market as more energy is needed. The demand for AI has also increased the amount of e-waste created each year, which is very harmful to the environment. These factors show how the electrical power infrastructure behind AI needs to be addressed, which is often overlooked due to its rising popularity and, in turn, demand.

Connection to STS

As AI is increasingly used across many new industries, demand for the data centers that power it grows. Because of this demand, many new data centers are being built, which require electrical power infrastructure. These new data centers are affecting the water and power grids of the communities around them. These stories are not discussed because they are overlooked on social media. This is because the hype around what AI can do is seen as perfect, with no consequences. The amount of research and development in AI has also skyrocketed in the past few years. This is an example of the correlation between the growing demand for a new technology and the disregard for its potential side effects on society and the environment.

Information on the topic

AI Power consumption

Using AI to find information, create images, or perform any action requires computational power. These computations take place in data centers scattered throughout the world. The largest of these is China’s Telecom Inner Mongolia Information Park, which is just over 10 million square feet. This power adds up when millions of people use these tools every day. To quantify the growing power consumption of these data centers, scientists estimated that the power required for data centers in North America increased from 2688  megawatts in 2022 to 5341 megawatts in 2023, driven by the rise of

mainstream generative AI (Zewe 2025). The global power consumption of these data centers totaled 460 terawatt-hours in 2022, making them the 11th-largest electricity consumers in the world (Zewe 2025). These numbers show how the rise of mainstream generative AI, such as ChatGPT, has affected the global power infrastructure. If this trend continues, AI/data centers will become among the world’s largest power consumers. With many new industries advocating for its integration into almost everything, this seems inevitable. This will not be sustainable with today’s electrical power technology. The majority of people who use these tools in their daily lives have no idea of the impact they have on the environment, largely because of the mainstream media.

AI E-waste

Another issue with the growing demand for AI and, in turn, the use of data centers is e-waste. This type of waste comes from hardware components, such as the processors and memory used in AI. This waste contains harmful materials such as mercury and lead (UN 2025). These materials can leak into the soil and groundwater around dump sites. This is toxic to both the surrounding environment and people. Many of the main components used in the hardware that drives AI use rare earth metals, which are harder to find and extract. It is estimated that making a 2-kilogram computer requires 800 kilograms of raw materials (UN 2025). In 2021, the total amount of e-waste was 57.4 million tons, and experts estimated that 347 tons of non-recycled e-waste remained by the end of 2024 (Khattak 2024). If this waste is not addressed as the demand for AI continues to grow, it could destroy whole communities and ecosystems.

Missing voices

The communities surrounding data centers and e-waste facilities are suffering from their operations. Data centers require a massive amount of water to cool their hardware components. Globally, AI is expected to use 4.2-6.6 billion cubic meters of water by 2027 (UN 2025). This puts stress on the local water reserves around data centers. As mentioned before, e-waste from these data centers can leak into surrounding groundwater, killing wildlife and tainting drinking water. According to the World Health Organization, exposure to e-waste can cause premature birth, changes in lung function, and respiratory issues (Khattak 2024). The growing demand for new data centers and the construction of new data centers have also put strain on the GPU and memory markets. Manufacturers are prioritizing the production of high-end memory for data centers over consumer products, driving prices up drastically (Bajarin 2026). These issues are not brought up because the media makes AI seem like the solution to everything.

Conclusion

Exploring the effects AI has on the global energy market, energy consumption, and its e-waste production helps create an understanding of issues with current global electrical power infrastructures. Understanding these issues can help create a plan for what needs to change or be engineered to keep up with the growing global power consumption. Finding better ways to predict these issues before they become a problem can reduce their impact on electrical power infrastructure. Acknowledging missing voices in engineering is also important for understanding how past and future inventions affect communities and society as a whole.

References

https://www.chintglobal.com/content/dam/chintsite/global/en/about-us/news-center/news/Data%20Center%20in%20Wuzhen%2020241203-9.jpg

by Unknown Author is in the public domain.

Schneider Electric. (June 12, 2023). Artificial intelligence power consumption and share of total data center consumption worldwide in 2023, with forecasts to 2028 [Graph]. In Statista. Retrieved February 18, 2026, from https://www-statista-com.libproxy.clemson.edu/statistics/1536969/ai-electricity-consumption-worldwide/

Zewe, A. (2025, January 17). Explained: Generative AI’s environmental impact. MIT News; Massachusetts Institute of Technology. https://news.mit.edu/2025/explained-generative-ai-environmental-impact-0117

United Nations. (2025, April 7). Artificial intelligence: How much energy does AI use? United Nations Western Europe. https://unric.org/en/artificial-intelligence-how-much-energy-does-ai-use/

Environmental impact of AI – Big AI’s dirty secret – IMD. (2025, March 20). IMD Business School for Management and Leadership Courses. https://www.imd.org/news/artificial-intelligence/big-ais-dirty-secret-the-environmental-cost-of-generative-ai/

Khattak, R. (2024, October 14). The Environmental Impact of E-Waste. Earth.org.

O’Donnell, J., & Crownhart, C. (2025, May 20). We did the math on AI’s energy footprint. Here’s the

story you haven’t heard. MIT Technology Review.

https://www.technologyreview.com/2025/05/20/1116327/ai-energy-usage-climate-footprint-big-te

ch/

Bajarin, T. (2026, January 16). As AI Eats Up The World’s Chips, Memory Prices Take The Hit. Forbes.

https://www.forbes.com/sites/timbajarin/2026/01/16/as-ai-eats-up-the-worlds-chips-memory-price

s-take-the-hit/

AI Use Acknowledgements

I used ChatGPT and Grammarly to help me find information about AI power consumption that fits the goals of this textbook chapter. It gave me sources from Google and Google Scholar. I then used the other sources I had gathered to apply other elements to the chapter.

OpenAI. (2026). ChatGPT [Large language model]. https://chat.openai.com/chat

Grammarly. (2026). Grammarly AI [Generative AI tool] https://www.grammarly.com