The Trump administration has formally repealed the Biden-era Mercury and Air Toxics Standards (MATS), rolling back restrictions on mercury and other hazardous pollutants from power plants. This regulatory reversal arrives at a critical moment, as electricity demand in the United States is experiencing a significant uptick, driven largely by the rapid buildout of new data centers to power artificial intelligence. The move is particularly consequential for coal-fired power plants, which are major sources of the pollutants now facing fewer constraints.

The Mercury and Air Toxics Standards, first established in 2012 and later strengthened, were designed to limit emissions of mercury, arsenic, and other toxic metals from coal and oil-fired power plants. The Biden administration had sought to tighten these rules further, citing public health benefits. The repeal by the Trump Environmental Protection Agency represents a major policy shift, prioritizing reduced regulatory burdens on the energy sector over established environmental and health safeguards. The timing is inextricably linked to the national energy landscape, where the explosive growth of compute-intensive AI models is creating unprecedented demand for reliable, baseload power.

This policy change directly impacts the energy infrastructure that supports the AI industry. Data centers, especially those training and running large language models, require massive, constant electricity. While many tech companies have committed to powering these operations with renewable energy, the reality of grid capacity often leads to reliance on existing fossil fuel plants, including coal, to meet immediate demand. The repeal of MATS lowers the operational costs and compliance hurdles for these plants, potentially incentivizing their continued or even expanded use to feed the power-hungry AI boom. This creates a tangible tension between technological advancement and environmental responsibility.

Analysts warn that the confluence of relaxed pollution controls and rising energy consumption could lead to a measurable increase in harmful emissions. Mercury is a potent neurotoxin that can accumulate in the food chain, posing risks to human health, particularly for children and pregnant women. Other pollutants covered under the standards are linked to respiratory illnesses and heart disease. The AI industry's growth, therefore, is now indirectly tied to a potential regression in air quality and public health outcomes, a trade-off that has sparked concern among environmental groups and public health advocates.

The situation underscores a broader, looming challenge for the tech sector: the environmental footprint of artificial intelligence. As companies race to develop more powerful models, the energy and infrastructure requirements become central to both their operational costs and their ESG (Environmental, Social, and Governance) profiles. The regulatory environment plays a decisive role in shaping the sustainability of this growth. The repeal of MATS may offer short-term grid stability and lower energy costs for data center operators, but it externalizes the long-term health and environmental costs onto the public, setting a precedent that could define the environmental impact of the AI era.