Marginalized communities all around the world are facing increased pollution and environmental injustice as a consequence of advancing tech like Generative AI. These communities became dumping grounds as a direct result of environmental racism, the act of intentionally building facilities that produce pollution and waste in areas where marginalized communities reside.
In 2014, there was the water crisis in Flint, Michigan, where lead pipes contaminated drinking water. In Norway, there was a three-year-long dispute over the country installing the country’s largest windfarm on the indigenous Sami people’s land, and the entire Global South faces exploitation and environmental degradation, disguised as green energy. Now, there is Boxtown in Memphis, Tennessee.
Water tower in Flint, Michigan
Generative AI Creations Don’t Come Out of Thin Air
The National Association for the Advancement of Colored People (NAACP) has filed a lawsuit against xAI, Elon Musk’s artificial intelligence company. The lawsuit alleges that xAI and Musk have participated in gas pollution after hastily setting up two data centers powered by gas turbines in the mainly black community of Boxtown. On top of that, the NAACP claims xAI built these turbines without a permit or pollution control.
xAI has become the biggest contributor of smog pollution in Shelby County. Even before the implementation of their data centers, Memphis, a predominantly black city in Shelby County, was ranked poorly in air pollution. The city is the asthma capital of the US, never receiving higher than a D ranking in air quality.
Would You Trade Your and Your Family’s Health for a Job?
While AI data centers bring in new jobs to these areas, many aren’t directly related to the data centers themselves. According to Business Insider, of the 4.7 million jobs brought in by these data centers, only around 600,000 are directly related to work in the data center. Most of the job creation involves the construction and maintenance of the data centers or indirect work. Creation of new jobs is beneficial in lower-income areas, but when the data centers are projected to increase energy demand by 29% or even a whopping 166% from 2023 to 2030, the risk does not seem worth the reward. This is not even to mention the fact that AI data centers use around 5 million gallons of water to run, as well as the health concerns residents will face from increased pollution.
Sources of global electricity generation for data centers, 2020-2035
Data via IEA, CC BY 4.0
Tech Advancing So Fast Regulations Fail to Keep Up
MIT’s paper further echoes the environmental damages Gen-AI will bring if checks and balances aren’t put in place. Just increasing efficiency isn’t enough. Leaders of Gen-AI companies will have to play a major role in steering Gen-AI development in a direction that also focuses on reducing AI’s economic, environmental, and social harm, and their stakeholders must hold them to it. Furthermore, people need to be aware of the environmental racism currently happening around them. Without awareness of how these technologies impact marginalized communities, there can be no accountability. Tech is going to advance and need energy, but we don’t have to allow companies to use any land they please as a dumping ground.
Projected electricity use rise of data centers and AI through 2030
Credit: TPIN via Frontier Group
Everyone is welcome! We offer DEI programs tailored to meet the needs of all experience levels—find out more about our workshops here.
