Trump Seeks Looser Clean Water Act Rules for Big Tech Data Centers
The White House has unveiled a new “AI Action Plan” that proposes significant changes to environmental regulations, particularly those under the Clean Water Act, to accelerate the development of artificial intelligence (AI) infrastructure, including data centers. These proposed changes align with suggestions previously made by major tech companies like Meta and lobbying groups representing firms such as Google and Amazon Web Services.
The core of the White House’s plan involves streamlining permitting processes and potentially creating new categorical exclusions from environmental review for AI-related data centers. Specifically, the plan aims to explore the need for a nationwide Clean Water Act Section 404 permit for data centers, which would remove pre-construction delays and limit public participation and federal review. Section 404 of the Clean Water Act regulates the discharge of dredged or fill material into “waters of the United States,” including wetlands, and typically requires a permit for activities like filling marshland or redirecting streams. Obtaining individual 404 permits can be time-consuming and costly.
This push for deregulation comes amidst growing concerns about the environmental footprint of AI and data centers. Data centers, which are essential for training complex AI models and serving user requests, consume vast amounts of water primarily for cooling their equipment. On average, a data center can use approximately 300,000 gallons of water per day, equivalent to the water usage of 100,000 homes. Some estimates suggest a single large-scale data center can consume between 3-5 million gallons of water daily, comparable to the water usage of a town with 30,000-50,000 residents. By 2027, AI’s water usage is projected to reach an alarming 6.6 billion cubic meters globally. The water used for cooling is often fresh water, intensifying pressure on freshwater resources, especially in water-scarce regions. Furthermore, the heated water from cooling systems often cannot be released back into the environment due to the risk of thermal pollution.
The energy demands of AI data centers are also substantial, leading to increased electricity consumption and associated greenhouse gas emissions, particularly if powered by fossil fuels. The International Energy Agency estimates that global electricity consumption by data centers could more than double by 2030.
Environmental groups and advocates have voiced strong opposition to the proposed rollbacks, arguing that they would prioritize corporate interests over public health and environmental protection. Critics highlight that weakening these protections could lead to more polluted water, increased flood damage, and fewer safeguards for communities. They also point out that while tech companies often publicize ambitious climate goals, their lobbying efforts on climate policy have been less proactive.
The Trump administration has a history of environmental deregulation, having previously repealed the Clean Water Rule in 2019, which removed protections for a significant portion of the nation’s streams and wetlands. A 2020 rule, the Navigable Waters Protection Rule, further deregulated a quarter of all wetlands and a fifth of all streams in the United States, including 30 percent of watersheds supplying drinking water.
The White House’s “AI Action Plan” frames environmental permitting as an impediment to rapid infrastructure development, stating that the current system makes it “almost impossible to build this infrastructure in the United States with the speed that is required.” While the plan suggests using AI to expedite permitting processes, it also explicitly includes provisions to reduce or streamline environmental regulations under the Clean Water Act and Clean Air Act. This approach is seen by some as a “sellout” that could lead to increased electricity costs for everyday Americans to subsidize discounted power for massive AI data centers.