Illustration by Lexi Forman, Cartoonist.
Chapman student Ciarra Hawkins fears that one day, when people think of her hometown of Gilroy, California, they will no longer picture the sprawling garlic fields that earned the city its title as the Garlic Capital of the World.
Instead, she worries they will picture a massive data center, a symbol of the growing artificial intelligence industry and its increasingly visible environmental footprint.
Artificial intelligence can generate an essay in seconds, answer questions with a single prompt and help researchers analyze information that would otherwise take hours to process. But as the technology becomes increasingly integrated into everyday life, the infrastructure required to sustain it is expanding, raising questions about what communities like Gilroy stand to lose in the process.
AI systems require enormous amounts of computing power to operate, relying on data centers that consume significant amounts of electricity and water. As universities and technology companies continue to embrace AI, students and faculty are beginning to question whether its benefits outweigh its environmental consequences.
Artificial intelligence requires substantial computing power to train, operate and maintain. The data centers supporting these systems consume electricity and require water to cool the hardware that processes information.
According to the University of Michigan's Center for Sustainable Systems, AI and data centers contribute to growing energy and resource demands. MIT News also highlighted how generative AI's environmental impact extends beyond electricity consumption.
"Beyond electricity demands, a great deal of water is needed to cool the hardware used for training, deploying and fine-tuning generative AI models, which can strain municipal water supplies and disrupt local ecosystems," MIT researchers explained.
As demand for AI increases, so does the need for infrastructure capable of supporting it. However, the resources required to maintain that infrastructure are not unlimited, raising concerns about how data center development could affect surrounding communities.
Hawkins, a senior majoring in creative writing, said she is particularly concerned about the placement of data centers in agricultural communities.
Originally from Gilroy, California, Hawkins described growing concerns surrounding a proposed data center in her hometown. She said the project had faced community opposition and was put on hold amid residents' concerns. She recalled residents protesting and raising their concerns at city council meetings.
"I live in Gilroy, which is famously known as Garlicville," Hawkins said. "A lot of our economy is founded on agriculture. We're a big farming community."
For Hawkins, the concern is that the resources needed to support data centers could compete with those necessary to sustain local agriculture.
"We need the water, though, to grow the crops, to fund our city," she said.
Beyond water consumption, she also expressed concern about the electricity required to operate data centers, potential increases in utility costs and the broader effects on communities that depend on agriculture.
Her concerns reflect a larger environmental justice question: When technological development requires additional natural resources, who gets access to those resources, and who is expected to make sacrifices?
The expansion of AI infrastructure is not happening independently of other environmental pressures. Communities across the country are already navigating issues involving pollution, resource distribution and industrial development.
For Kyle Harp-Rushing, director of environmental studies at Chapman, the rapid adoption of AI in higher education reflects a larger transformation in how universities approach learning.
Harp-Rushing’s research has included examining environmental justice organizations in the Inland Empire, where warehouses, logistics facilities and particulate matter pollution have become significant concerns.
He described environmental justice as extending beyond reducing pollution. It also involves supporting communities facing overlapping social and environmental challenges.
When discussing AI, he expressed concern about society's rapid adoption of the technology without sufficient consideration of its consequences.
He described his outlook on AI in higher education as "pretty bleak," arguing that universities and other institutions are embracing the technology without adequately examining its broader effects.
Although his primary concern is the role of AI in education, he emphasized the importance of understanding the political and environmental systems that shape technological development.
His perspective raises questions about whether AI expansion constitutes meaningful progress when its benefits are separated from the environmental consequences borne by the communities that support it.
Not everyone believes that focusing exclusively on AI is the most effective way to address environmental concerns.
Joshua Dowd, a sophomore in computer science, acknowledged that AI is not environmentally harmless but argued that its impact should be considered alongside other major contributors to environmental degradation.
"If the environment is your main concern, you should be more concerned with a lot of things out there, not necessarily AI," Dowd said.
Dowd pointed to transportation, particularly car emissions, as another significant contributor to environmental damage. He suggested that society should consider broader energy solutions, including generating more electricity through alternative sources such as nuclear power.
However, he also acknowledged that AI's environmental footprint has been negative so far.
He explained that data centers' energy and water demands have created environmental costs, adding that he hopes the technology will eventually contribute to environmental solutions. So far, however, he has not seen that shift toward a net environmental benefit.
Dowd also raised concerns about how competition for water could affect different industries.
Agriculture, which typically accounts for significant water consumption, could face additional pressure as data centers require more resources. Depending on how water is allocated and priced, the consequences could extend to other users as well.
This creates a difficult balance between supporting technological development and maintaining access to the resources communities depend on.
For Hawkins, the environmental costs are reason enough to question whether the expansion of AI should continue without greater consideration of its consequences.
"I think it's really not good for the environment," Hawkins said.
Although she recognizes that data centers and AI may offer technological benefits, she questions whether those benefits justify placing additional pressure on agricultural communities.
The environmental debate also extends into universities, where students are increasingly encountering AI in their academic and professional lives.
For Harp-Rushing, the rapid adoption of AI reflects a broader shift toward treating higher education as a means of acquiring marketable skills rather than developing independent thought and understanding.
He said that students are under tremendous pressure to graduate with skills that will make them competitive in the workforce, contributing to the appeal of AI tools that promise to make academic work easier.
However, he questioned whether students are losing valuable learning experiences by relying on technology to complete tasks for them.
"What is the purpose of a higher education? What role should it serve in a society?" Harp-Rushing said.
He said that education should involve learning how to navigate uncertainty, develop independent perspectives and engage with difficult questions rather than immediately outsourcing the process to chatbots.
He also acknowledged that AI has applications in certain research fields, including environmental science, where machine learning can be used to work through complex problems.
However, he distinguished these applications from the everyday use of generative AI chatbots.
Joshua Fisher, an associate professor in Chapman’s Schmid College of Science and Technology, emphasized that AI should be taught as a tool rather than treated as a replacement for human understanding.
“AI should be properly taught as a tool in higher education, like the typewriter, calculator, and computer,” Fisher said. “But, like a calculator, students need to be taught to identify if the answer coming out is incorrect.”
The distinction is particularly important in environmental research. AI may help researchers analyze information and identify patterns, but students should not assume that technology alone can resolve the complicated political, economic and environmental problems surrounding climate change.
As artificial intelligence becomes increasingly embedded in everyday life, its environmental consequences raise questions that extend beyond individual use of technology.
For Harp-Rushing, the conversation is part of a larger ecological and political moment that requires society to reconsider what progress should look like.
He described a tension between reducing the potential harms of emerging technologies and imagining a better relationship between people, one another and the environment.
Rather than focusing exclusively on technological advancement, he encouraged people to think more intentionally about the relationships they build and the world they want to sustain.
Hawkins similarly believes that people should be more conscious of their AI-related decisions, particularly when those decisions have consequences beyond the classroom.
While the technology may offer convenience and new opportunities, its environmental footprint cannot be separated from the resources and communities that enable its development.
As universities, companies and consumers continue embracing artificial intelligence, the question is no longer simply what AI can do. Instead, it is whether the benefits of its expansion justify the environmental costs and who ultimately pays the price.