How the AI boom derailed clean‑air efforts in one of America's most polluted cities

The demand for electricity from the artificial intelligence boom and new data centers is causing the U.S. to reverse decades of environmental policy, particularly impacting efforts to reduce air pollution from coal plants. In cities like St. Louis, which already faces significant air quality issues, residents and activists, including Barbara Johnson of Metropolitan Congregations United, are expressing dismay over the rollback of tougher federal soot standards that were set to take effect in 2027. Initially, the Biden administration's standards would have required major polluters like Ameren's Labadie Energy Center, located west of St. Louis, to halve their soot emissions. However, the Trump administration scrapped these standards, citing the need to ensure grid reliability for surging data center demand. The U.S. Environmental Protection Agency (EPA) defended the rollbacks, while the U.S. Department of Energy (DOE) estimates AI and data center growth will add 50 gigawatts of electricity demand by 2030, nearly a 4% increase. St. Louis is projected to be one of the most impacted cities due to its existing poor air quality and the Labadie plant's high emissions of sulfur dioxide and nitrogen oxides, which impose an estimated $820 million in annual health costs on area residents. Clean air activists like Darnell Tingle of United Congregations of Metro-East and Patricia Schuba, who monitors coal plants, label the region a "sacrifice zone." Ameren stated the Labadie plant operates within current federal limits and will continue running for at least another decade to ensure reliable power. Meanwhile, data center developers, including Amazon Web Services, are planning large projects near St. Louis, such as a 1,000-acre development in Montgomery County, which will draw power from Ameren. The Data Center Coalition acknowledged its members' commitment to clean energy but emphasized that utilities, regulators, and policymakers are responsible for resource planning and generation decisions.