eWEEK content and product recommendations are editorially independent. We may make money when you click on links to our partners. Learn More.
A recent study by researchers at Caltech and the University of California Riverside reveals that AI’s massive power demands are creating dangerous air pollution that could lead to 1,300 premature deaths in the U.S. annually by 2030 and up to $20 billion in annual health-related costs.
Titled “The Unpaid Toll: Quantifying the Public Health Impact of AI,” the research paper highlights the understated environmental footprint behind AI advancements. As more data centers and computing facilities emerge to power AI systems, energy consumption spikes. This surge in electricity demand leads to increased emissions from power plants and diesel backup generators used to ensure consistent power supply. These emissions contains lung-penetrating fine particles smaller than 2.5 micrometers.
Disproportionate Impact on Vulnerable Communities
The researchers discovered that low-income communities near power plants and data centers face the highest health risks. Diesel generators commonly used as backup power sources emit 200 to 600 times more nitrogen dioxide than natural gas power plants, degrading air quality in these areas. However, the pollution doesn’t stop there. It drifts across state lines, impacting communities far removed from the original source.
“Pollution from backup generators at data centers in Northern Virginia drifts into Maryland, West Virginia, Pennsylvania, New York, New Jersey, Delaware, and the District of Columbia,” the study said, “creating regional public health costs of some $190 million to $260 million a year. If these backup generators emit at their maximum permitted level, the annual cost will become 10-fold and reach $1.9 billion to $2.6 billion.”
The Cost of AI-Powered Progress
The research paper stresses the mind-boggling scale of the issue. Training just one large language model, such as Meta’s Llama 3.1, generates air pollution equivalent to 10,000 round trips by car between Los Angeles and New York. By 2030, AI-related pollution could rival the emissions from California’s 35 million vehicles, surpassing even the steel industry’s impact on public health.
Shaolei Ren, an associate professor at UC Riverside and co-author of the study, highlighted a significant gap in corporate accountability.
“If you look at sustainability reports by tech companies, they only focus on carbon emissions and water use,” Ren said. “There’s absolutely no mention of harmful air pollutants, which are already creating a public health burden.”
The authors encourage AI companies to adopt more transparent practices by reporting carbon emissions and the air pollution resulting from their operations. They also recommend compensating affected communities and adopting sustainable energy practices to reduce environmental damage.
“If you have family members with asthma or other health conditions,” Ren said, “the air pollution from these data centers could be affecting them right now. It’s a public health issue we need to address urgently.”
Learn more about the energy consumption involved in powering AI and the effect it has on resources and the environment.