The EPA is drafting a rule change that would eliminate the requirement for states to collect public feedback before issuing air pollution permits for so-called 'minor source' facilities, a category that includes data centers, landfills, laundromats, and steel mills. The move comes as communities increasingly use public hearings to challenge AI data center construction over air quality and health concerns. Attorneys general from 14 states and three cities have signed a letter opposing the change, arguing that minor sources can still cause serious health and environmental harm.
Chris Malone, who led OpenAI's data center strategy after joining from Meta and Google in March 2024, has left the company. OpenAI confirmed it recently restructured its infrastructure organization, shifting oversight so that Malone no longer reported to President Greg Brockman but instead to VP Sachin Katti before his exit. The company says several other executives, including Uday Ruddarraju, Brent Mayo and Spas Lazarov, remain in charge of its data center efforts.
Governor Josh Shapiro has pursued a strategy aimed at attracting artificial intelligence investment and jobs to Pennsylvania while simultaneously working to shield electricity ratepayers and nearby communities from the costs and disruptions data centers can bring. The approach seeks to balance economic development ambitions with consumer protection and local concerns as AI infrastructure expands rapidly across the country.
Chris Malone, who led OpenAI's data center buildout since March 2025, has left the company, according to CNBC. Malone previously worked as a distinguished engineer on infrastructure at Meta and Google before joining OpenAI to help manage its roughly $600 billion compute spending plan through 2030. OpenAI says it has reorganized its infrastructure team and maintains strong leadership despite the exit.
Chris Malone, who led OpenAI's data center operations, has exited the company. His departure adds to a growing list of senior executives leaving OpenAI as it accelerates infrastructure spending and moves toward a possible IPO.
A new analysis of seven states housing roughly half of America's data centers finds that generating the electricity these facilities consume uses far more water than cooling their servers does. Across Virginia, Texas, California, Illinois, Georgia, Ohio and Arizona, data center power demand draws about 3.4 trillion gallons of freshwater a year—about 12 times the combined annual water use of Los Angeles, Phoenix and Washington, D.C. Most of that electricity comes from water-dependent power plants, and two-thirds of them sit in regions already facing significant water stress or drought.
A Congressional Research Service report cited by the Library of Congress estimates US data centers consumed 17 billion gallons of water in 2023, about triple the 2014 level. Over 80% of that footprint comes indirectly from electricity generation needed to power the facilities, rather than from onsite cooling systems.
Global Energy Monitor's updated tracking shows gas-fired power capacity in development specifically for data centers reached over 189 gigawatts by mid-2026, up from 97 gigawatts at the end of 2025 and just 4 gigawatts in early 2024. The surge reflects tech companies building private 'behind-the-meter' gas plants to power AI data centers rather than waiting for grid connections.
Flock Safety's network of AI-powered surveillance cameras has expanded rapidly across US cities and towns, but communities from both political sides are now organizing against it. Critics cite privacy concerns, lack of oversight, and fears about how police and other agencies use the collected data.
Rising electricity demand from AI data centers is spurring investment in solid-state power transformers, which use semiconductor switching instead of the century-old copper-coil design of conventional transformers. Unlike traditional units, which are hand-built and face multi-year delivery backlogs, solid-state versions can be mass-produced using materials like silicon carbide, making them smaller, lighter, and modular. Tech companies building AI infrastructure are especially drawn to them because they suit the direct current power setups increasingly used to run dense racks of AI chips.
In an interview with Michael Cohen, President Trump argued that towns rejecting data centers are making a mistake, claiming the facilities bring jobs and 'tremendous tax revenue.' His remarks come as data centers face bipartisan public backlash over energy use, water consumption, and local disruption.
Reporting summarized by Futurism shows AI industry leaders responding to growing public frustration over AI products and data centers not by addressing underlying harms, but by launching messaging campaigns framing the backlash as a 'trust' problem. Mark Zuckerberg published a lengthy essay defending AI's promise and criticizing industry pessimists, even as Meta's own internal practices reportedly contradict his optimistic claims.
GoKawiil