AI Usage Sparks New Environmental Concerns

As technology companies invest in artificial intelligence, they leave their carbon neutrality goals behind.

The technological advancements of AI cannot seem to keep up with its environmental impacts. Photo Credit: ThisIsEngineering from Pexels

Immense artificial intelligence investment is forcing major technology companies to push back on their carbon neutrality goals. 

As of its July 2024 environmental report, Google no longer claims to be carbon neutral and now aspires to be carbon-free by 2030. Similarly, despite ambitions to be carbon-negative by 2030, Microsoft’s indirect emissions have grown by 30%, according to its May report.

Along with other companies, like Meta, they credit the increased use of AI as the reasoning behind new data centers needed to meet rising energy needs. These announcements have prompted questions about the impacts of AI on the environment.

For instance, AI requires significantly more energy to generate a response than a normal online search, and heat output from servers also increases the energy consumption of data centers.

“When you’re doing an AI search, those servers run a lot hotter than traditional Internet searches, and that requires additional cooling from HVAC units or water,” said computer science major Ellie Knapp ‘26. “That just means that globally, as AI has become more accessible and more popular, a lot more HVAC and water are needed to cool the servers, and that has an environmental cost.”

Knapp worked as a tech intern with the sustainability team at Comcast this summer, and through her internship, she measured metrics of AI energy consumption to help maximize efficiency.

“It’s in the development of the AI that so much energy is used because AI requires a vast amount of data to do pattern recognition,” Knapp said. “And that’s where it can be concerning, is [that there are] so many separate developments of AI because everyone wants to have their own competing AI functionality.”

This growing demand for energy threatens efforts to move to more sustainable energy sources. According to The Washington Post, the rising need for power has slowed efforts to close coal power plants in Kansas, Nebraska, Wisconsin and South Carolina.

“AI is one of the factors that slows down progress in reducing emissions because the vast majority of our electricity is generated by burning methane,” said Richard Feldman, chair of the Marist College environmental science and policy department and an associate professor.

The International Energy Agency estimates that AI growth will push data center energy consumption up to 6% of the total electricity usage in the U.S. by 2026. With fossil fuels still having produced 83% of the nation’s energy in 2023, according to the U.S. Energy Information Administration, AI’s energy consumption will directly feed carbon dioxide emissions. However, the decreasing price of renewable energy sources like solar panels could help alleviate some of the long-term problems.

“There needs to be a much more concerted effort by Google and everybody in this business to be powering their facilities with sustainably produced electricity,” Feldman said.

Additionally, there are ways to make AI more energy efficient through code and algorithms. Knapp predicts that the financial incentives will encourage technology companies to pursue these methods. 

“This is not just an environmental cost for these companies, this is money they’re spending on energy. It’s going to force companies to get more creative and to design their AI in more ethical ways because it’s not sustainable financially,” Knapp said.

AI also offers advantages in other areas of environmental science, like tracking both pollution and endangered species. AI’s ability to process large amounts of information can help compensate for understaffed fields. 

“Those two areas of environmental science don’t have enough people or funding to really get the job done to the level it needs to be,” said Feldman. “So, viewing AI as a powerful tool to assist those people, that is where my hopefulness is.”

With such a novel technology, the long-term impacts of AI on the environment are yet to be determined. Though the majority of these impacts will be defined by how major technology companies respond, individual consumers can still be mindful. 

“I highly encourage people to be much more intentional with their interactions with AI tools,” said Knapp. “Instead of spamming it with multiple requests, just be more thoughtful, so that you’re minimizing your number of searches.”