In a March 2025 Bark survey, 30 percent of students reported using ChatGPT once or twice a week, and 27 percent of students reported using the chatbot several times a day. Students today use generative artificial intelligence (AI) to rewrite college essays, solve tricky math problems even Photomath can’t answer and to help formulate the perfect email, but few are aware of the environmental cost of this suspiciously flawless platform.

How AI consumes more energy than other technologies: 

When most people think of ChatGPT or other technologies, they think of them as being entirely removed from the real world. However, the infrastructure required to store all this information goes beyond the “cloud.” Indeed, the physical data centers designed for holding this information consume massive amounts of energy. Google itself revealed that just one Google search generates 0.2 grams of carbon dioxide emissions — not to mention the amount of energy required to power the data centers and the entire energy generation process. But generative AI takes energy waste to a whole new level. As AI formulates the perfect responses using information from across the web, the environmental impact of ChatGPT goes far beyond a typical Google search.

For every search on ChatGPT, the massive language model is estimated to use 2.9 Wh of energy, about 10 times what it takes for a single Google search. Annual energy consumption for ChatGPT is projected to reach 226.8 gigawatt-hours, an amount that could power 7.9 million iPhone 15s every day for a year, according to RW Digital. Shaolei Ren, a professor at the University of California (UC), Riverside, who has researched the large environmental footprint of AI, explained why chatbots in particular have such a large impact on the climate. 

In general, generative AI tends to use more resources than other types of AI models because it has more parameters… More parameters generally means you’re going to have more resource consumption, more energy, more environmental impacts and more water and air pollution,” Ren said. 

AI’s role in the classroom: 

All of this comes as AI becomes increasingly more prevalent in the classroom. Senior and prospective environmental science major Rumi Rezoni commented on the increased AI use she has seen in classes. 

“[ChatGPT] is not a necessity. People just use it for everything because they don’t want to think critically anymore. We’ve survived thus far without the use of AI, and it shouldn’t be hard to stop using it, but students are using it so much that it becomes difficult for them to become less reliant,” Rezoni said.

Increased AI usage makes understanding its environmental impact all the more important. Ren anticipates this rapidly growing industry will have serious environmental impacts in the future.   

“Generative AI is being used more and more, and it’s growing really fast compared to other sectors. [According to] the Berkeley lab report for 2028, the additional energy that we will use for data centers, compared to what we are using now, is driven almost entirely by generative AI,” Ren said. “We should pay particular attention to generative AI because the decisions that we make today, like where to build the data centers, will have impacts for many years to come.”

The conservationist approach: 

This is not to say that we should stop our use of generative AI altogether. As more and more platforms integrate generative AI into the very makeup of their interface, as with the new Google Gemini feature, it seems that this technology will only become more difficult to escape. Additionally, environmental costs are often a natural result of technological developments, like the air pollution that accompanied the development of automobiles. Instead of writing off this technology altogether, many experts, like Sustainable Agriculture teacher Joe Stewart, advocate for finding a balanced relationship with the technology. 

“We want to have conservation, but we have to look at the political, social and economic reality. If we just don’t use AI, we’re going to put ourselves at a disadvantage in other regards. The answer, to me, is [thinking about,] ‘What are the ways that we can use AI when we need to, and not use it when we don’t?’” Stewart said.

This conservationist approach is one that Rezoni champions as well.

“The only way we can make people actually take steps in their daily lives to combat climate change is [through education]. I think we should really try to teach [about environmentalism] in school, separate it from politics and bring it to the public as an imperative thing we must do [that is] essential for survival,” Rezoni said.

Responsibilities of tech companies:

Moving forward, it is crucial to recognize that the burden of AI’s climate impact is not solely the fault of the consumer. Tech companies have stayed remarkably silent about the environmental costs of the AI they incorporate into every product, leaving consumers in the dark about the impacts of their own carbon footprint. Consumers like Rezoni believe that transparency from tech companies is key to making substantial changes in consumer behavior. 

“Tech companies report every year everything that they made in profit. And I think that sustainability should also be reported — how much net carbon they introduce into the atmosphere and [how it’s impacting the environment],” Rezoni said.

Additionally, experts have noticed the ways in which tech companies have been selective in what information they release about the environmental impacts of AI.

“These companies should be more transparent in order for them to earn trust from the public. They are reporting their carbon pretty comprehensively. But most companies only focus on direct water consumption,” Ren said. “Companies don’t say anything about [air pollutants produced through their technology] and I think they should. This type of air pollutants have a direct and tangible impact on people’s health because they can travel hundreds of miles.”

How should AI be used in the future?

Stewart explained how the long-term impacts of AI should be weighed against the short-term benefits of the technology.

“We need to be as aware as we can and thoughtful about [using generative AI]. But ultimately, we as humans want to have health. We want to have freedom. So how do we have health and freedom in a long-term way?” Stewart said.

AI isn’t going away anytime soon. That being said, it is of the utmost importance that we use it in moderation and ensure that we only reap the benefits of such technology while limiting our environmental footprint. Ren advocates for both tech companies and consumers to assume responsibility and encourages the development of this technology in a sustainable, long-term fashion.  

“This study on the environment is not trying to stop this technology. It’s trying to advance this technology more responsibly to preserve our natural resources and make sure we are not abusing [them,]” Ren said.