The Trump administration this week released a plan to fast-track the growth of artificial intelligence (AI) in the US, delighting tech groups while alarming environmental advocates who point to the industry’s toxic emissions, high water usage and heavy reliance on fossil fuels.

The “AI action plan,” released July 23 by the White House, calls for the development of new AI data centers – huge facilities that house AI computing infrastructure – to be waived from typical assessment requirements under the National Environmental Policy Act, which determine a project’s environmental impact.

The plan also proposes expediting environmental permitting for such data centers by streamlining or reducing regulations under the Clean Air Act, the Clean Water Act and other environmental laws, and calls on agencies to make federal lands available for constructing data centers and their power generation infrastructure.

In accordance with the Trump Administration’s AI plan, the US Department of Energy today announced four sites across the country selected to invite private sector partners to develop AI data centers and energy generation projects – a “bold step” that will “accelerate the next Manhattan Project,” said Energy Secretary Chris Wright in a statement.

The plan is “nothing more than a thinly veiled invitation for the fossil fuel and corporate water industries to ramp up their exploitation of our environment and natural resources – all at the expense of everyday people,” Mitch Jones, the managing director of policy and litigation for the nonprofit Food & Water Watch, said in a statement. “The expanding data center industry is being leveraged as an excuse to prolong the life of filthy, climate-killing fossil fuel power and dangerous nuclear plants, and even build new ones.”

The document states that the US is locked in a “race to achieve global dominance in artificial intelligence (AI)” and that “Whoever has the largest AI ecosystem will set global AI standards and reap broad economic and military benefits. America’s environmental permitting system and other regulations make it almost impossible to build this infrastructure in the United States with the speed that is required.”

Doug Kelly, CEO of the Meta (formerly Facebook)-funded tech advocacy group the American Edge Project, said in a statement that the plan is “a giant leap forward in the race to secure American leadership in artificial intelligence.”

President Donald Trump on July 23 also signed three executive orders related to AI development, speeding up federal permitting of data center projects as outlined in the plan, establishing an AI exports program and preventing “woke AI” in the federal government.

Eating up energy

AI technology is more energy-intensive than many users may realize, with a typical search using OpenAI’s ChatGPT using almost 10 times more electricity than a Google search and ChatGPT’s daily electricity usage equaling that of 180,000 US households in 2024, according to Food & Water Watch. Each day, one Meta data center uses as much power as 7 million laptops running for a full workday, according to the group’s site.

There are currently more than 1,000 so-called “hyperscale” data centers worldwide, with over half of these massive facilities located in the US.

A recent report by the group suggests the AI’s industry’s impacts on water and energy resources could skyrocket in coming years, finding that by 2028 the energy demand from AI servers and data centers in the US could increase threefold. By that year, a million Olympic swimming pools of water may be needed to cool AI servers each year and the industry may require enough electricity to power more than 28 million American homes.

“Across the South we are seeing states and communities clamor for more information about the infrastructure demands for AI, not less,” Alys Campaigne, the climate initiative leader for the Southern Environmental Law Center (SELC), said in a statement responding to Trump’s AI action plan.

Creating an exclusion from NEPA would make it harder for impacted communities to get information about the environmental or public health impacts of new data centers, and the plan fails to mention the need to consult with communities beforehand, says the SELC press release.

Treated “like obstacles”

Training one AI model can produce as much air pollution as more than 10,000 car trips from Los Angeles to New York City, according to a study published on a preprint server in December 2024. The resulting public health burden of data centers may amount to over $20 billion per year in 2030 – double the health burden of coal-based steel production, the authors concluded, with an estimated 600,000 people suffering asthma symptoms and about 1,300 premature deaths resulting from the pollution.

Some US communities are already suffering alleged injustices from AI data centers. In a predominately Black community in Memphis, a supercomputer facility owned by xAI, a company founded by tech billionaire Elon Musk, has reportedly increased area smog by 30-60% as it releases toxic pollutants. Dozens of methane gas turbines owned by xAI are unpermitted, according to SELC.

The National Association for the Advancement of Colored People (NAACP) in June sent the company a letter threatening to sue if it continues using unpermitted gas turbines at its data center.

“All too often, big corporations like xAI treat our communities and families like obstacles to be pushed aside,” NAACP President and CEO Derrick Johnson said in a statement. “We cannot afford to normalize this kind of environmental injustice.”

(Featured image by Alex Shuper via Unsplash.)