With the recent rise of artificial intelligence (AI), society has begun to use AI chatbots for nearly anything. Whether writing a school essay, finding recipe inspiration, or asking for advice, AI is everywhere. While AI seems like an easy solution to many questions, it also comes with high environmental costs. AI data centers’ use of electricity and water results in severe carbon emissions and excessive consumption of natural resources, which are and will continue to be detrimental to the environment. 

Currently, the U.S. is the leading country in AI development with the most data centers, followed by China and Germany: Data Center Map estimates that there are about 4000 AI data centers in the U.S. Each of these centers contains several high-density computers called Graphic Processing Units (GPU), which take part in the training of artificial intelligence, analyze large quantities of data, and are what make the chatbots able to answer your everyday questions. Data centers have existed since before AI, but the increased use of these facilities for AI has driven the rise in their climate impact. 

Professor Christine Hoffman at Dominican University of California has her PhD in applied mathematics and works with students to train and apply AI models. She noted that the process of training the model is what consumes the most energy and resources. As AI models get larger and more complex, the amount of energy and resources it takes to train them increases. 

“The training process, when you’re giving it hundreds of thousands of images, for example, takes a lot of energy and a lot of computational resources. When I train an AI model, and I’m using relatively small models, they can still take days to train. Training these models over time definitely requires a large amount of resources because that computer has to be on and running day and night for days,” Hoffman said. 

As these data centers operate, they require millions of gallons of water to cool the data servers and prevent overheating. According to the Environmental and Energy Study Institute, data centers in the U.S. used around 211 billion gallons of water to cool computers in 2023, and this number is only projected to increase. The use of water for AI is approaching the annual consumption of bottled water across the globe. 

While the use of natural resources by artificial intelligence is significant, the greenhouse emissions may also have substantial environmental impacts. With the rise in AI, several large companies’ previously announced emissions-reduction goals have not progressed as planned. Companies such as Google and Microsoft, which had certain pledges of zero emissions, have faced setbacks while trying to compete in the expanding AI field. 

The carbon emissions from AI databases alone are predicted to be between 32.6 and 79.7 million tons of carbon dioxide in 2025, according to an article from Science Direct. The same article states that the amount of emissions from AI data centers is similar to the emissions from all of New York City. While the environmental impacts of AI seem daunting already, the continuing expansion of AI will only increase these impacts. 

Archie Williams Social and Environmental Academy Dedicated to Improving School and Community (SEA-DISC) Academy and AP Environmental Science teacher Clarke Bugbee teaches students about how human activities affect the environment. He believes that one of the hardest things about addressing the environmental impact of AI is that it isn’t widely discussed.

“It’s so tempting for people to use AI, and of course, people are using it without knowing about its impact, and nobody thinks about the environmental issue because it’s not really put out there much. So how can they think about it, right? Most people just don’t know at all,” Bugbee said.

Within SEA-DISC, certain groups of students chose to learn about how AI affects the environment for a group project. Archie Williams junior Hayden Bremer is a part of SEA-DISC and is working on this research project. 

“We’re learning about how the use of AI affects energy costs for local communities, and how much water consumption it uses daily and yearly. It’s really started making me reflect on my own usage of AI,” Hayden said.

While the current effects of artificial intelligence are intimidating, AI is continuing to grow. On Jan. 21, 2025, President Donald Trump launched the Stargate Initiative, a 500-billion-dollar project looking to build more databases and upgrade AI infrastructure across the world by 2029. While the idea behind increasing the use of AI is with the hope that it can help to find a cure for cancer, among other things, the world may be facing a more pressing issue of climate change before then. 

Currently, AI is highly unregulated, particularly regarding its environmental impact. Limited information from AI databases has made it difficult to understand the full effects artificial intelligence has on the Earth’s natural resources and carbon emissions. Laws have been proposed to request that databases provide more accurate numbers or study AI’s impacts, but none have been passed yet. 

Bugbee argues that, without regulations, it is difficult to get society to care about the effects of their AI use on the environment when they can’t directly see them. AI assistance is automatically enabled on devices and makes everyday tasks easier, so it’s difficult to get users to go out of their way to stop using it. 

“It’s interesting, you know, you can’t get people to recycle, right? We have all the stuff, yet we still can’t get people to recycle. If we can’t do that, how do you think we’re going to get people to discuss and to think about their use of AI? It’s on our phones with Google, and much of it is automatically done. Most of the time, you have to go out of your way to opt out of certain AI tools. It’s going to be very hard to get people to be conscious about it; people really love [AI],” Bugbee said. 

The convenience of AI, along with the overall lack of knowledge about its environmental impact, makes it potentially harmful if used too frequently. While there are potential positives to AI, such as solving large world issues, the damage that it is causing may outweigh its usefulness. As AI becomes more popular and the market becomes more competitive, stricter regulations and more transparent data will be imperative to protecting the environment.