President Donald Trump is weakening environmental regulations to accelerate the construction of AI data centers, a move that former EPA officials warn raises health risks for Americans, according to a new report and briefing covered by The Verge and other outlets. The group of agency alumni is urging the administration to adopt a “Data Center Health Protection Pledge” to signal it is taking the environmental threat seriously, though they acknowledge the plea may be futile.
The core allegation centers on a policy shift initiated at the start of Trump's current term. His EPA administrator, Lee Zeldin, explicitly proclaimed a goal of making America “the AI capital of the world” through deregulation. Since that declaration, Zeldin and the White House have rolled back dozens of environmental rules while simultaneously cutting staff and funding at the EPA, systematically reducing its capacity to monitor and enforce remaining protections.
The former officials argue that this deregulatory spree, combined with the administration's broader attacks on renewable energy incentives, has created a permissive environment for the explosive growth of AI data centers. The consequence, they state, is that this new wave of infrastructure is "increasingly powered by dirty energy." This shift in power sourcing carries significant health implications that extend beyond local communities. The officials emphasize that pollution from these energy sources can impact Americans' health even if they do not live near a data center, as emissions contribute to broader regional air quality issues and climate change.
Their warning carries a particular sting because it comes from individuals who helped run the agency now overseeing these changes. They frame the administration's actions as a betrayal of the EPA's foundational mission, which was created to "protect human health and the environment." The current trajectory, in their view, prioritizes unfettered industrial expansion—specifically in the strategically important AI sector—over these core public health responsibilities.
The context for this conflict is the unprecedented demand for computational power driven by the generative AI boom. Training and operating large language models like those from OpenAI, Anthropic, and Google require massive, energy-intensive data centers. The industry's growth has sparked a race to build new facilities at a breakneck pace, often straining local power grids and leading utility companies to delay retirement plans for coal-fired power plants or propose new natural gas plants to meet the demand.
The Trump administration's approach appears designed to remove perceived regulatory obstacles to this construction, framing environmental reviews, emissions controls, and clean energy mandates as bureaucratic red tape slowing down American innovation and competitiveness. This aligns with a long-standing political philosophy that views economic growth and deregulation as intrinsically linked. However, the former EPA officials' report positions this not as an economic necessity but as a direct trade-off, where speed in the AI race is being purchased with increased public health risk.
The call for a "Data Center Health Protection Pledge" is a symbolic attempt to challenge this trade-off. It asks the administration to publicly commit to ensuring that the AI infrastructure boom does not come at the expense of air quality and health. The fact that the officials themselves see the request as "perhaps futile" underscores the entrenched nature of the policy direction. It suggests they view the deregulation as a deliberate, ideological choice rather than an oversight, and one that is unlikely to be reversed through appeals to the agency's original mission.
This story highlights a critical, growing tension at the intersection of technology policy, industrial strategy, and environmental regulation. As AI becomes a central plank of national economic and security strategy, the environmental footprint of its underlying infrastructure is becoming impossible to ignore. The debate is no longer just about carbon emissions and climate goals, but about immediate, localized health impacts from pollutants like particulate matter, nitrogen oxides, and sulfur dioxide associated with "dirty energy" sources.
The warning from the EPA alumni serves as a sharp critique that the current U.S. path to AI dominance may be replicating the environmental mistakes of past industrial revolutions, despite available cleaner alternatives. It frames the administration's policies as a choice to grant the data center industry a regulatory "pass" to pollute, prioritizing short-term geopolitical gains in the AI race over long-term public health and environmental stewardship. The report sets the stage for increased scrutiny of the energy procurement practices of major tech companies building these centers and the regulatory frameworks that govern them, making the environmental cost of artificial intelligence a central policy battleground.
