In the global race to lead on artificial intelligence, President Donald Trump’s administration has made its goal unmistakable: the U.S. should become the world’s AI powerhouse.
To get there, the Environmental Protection Agency has taken on an unexpected role. Instead of focusing first and foremost on protecting people and the environment, the EPA is increasingly focused on clearing the regulatory path for rapid growth in data centers, chip factories and other AI-related infrastructure.
That shift has created a striking imbalance.
The agency is sprinting in helping companies build the physical backbone of the AI economy but dragging its feet when it comes to using those same powerful tools to improve environmental protection and public health.
Under Administrator Lee Zeldin, the EPA has positioned itself as a facilitator for the tech sector and the fossil fuel industry. Reviews of new chemicals are rushed, hurdles lowered for air pollution permits and approvals streamlined, all under the pretext of keeping the U.S. competitive.
The emphasis is clear: move fast.
This deregulatory zeal stands in stark contrast to the agency’s own lethargic embrace of AI to actually improve environmental and public health outcomes for communities across America.
So far, the EPA’s most visible efforts focus on basic office functions, such as sorting and summarizing public comments. These tools may save staff time, but they barely scratch the surface of what AI could do for the agency’s core work.
The same technologies that power large language models could help flag dangerous chemicals earlier, identify pollution hotspots more precisely or focus inspections where problems are most likely to occur. Instead, AI is largely being treated as a productivity aid, not a mission-critical tool.
This isn’t just a missed opportunity — it’s a growing risk. Companies across the economy are rapidly adopting AI to improve efficiency, predict failures and optimize complex systems. If the EPA doesn’t keep pace, it risks becoming a referee who no longer understands the speed of the game. In a fast-moving field, that kind of mismatch matters.
One way forward can be found in the agency’s own past.
Three decades ago, the EPA launched Project XL, an initiative that encouraged companies to try new approaches, as long as they could demonstrate better environmental results. Instead of following rigid rules, participating companies were allowed to experiment with innovative methods that delivered stronger outcomes. The idea was simple: reward better performance, not just box-checking. It was meant to be a new paradigm for environmental regulation and provide a way of moving state-of-the-art technology into the mainstream.
That effort ultimately faded, weighed down by complexity, high costs and concerns about oversight. But the core idea remains sound — and may be even more relevant in the age of AI.
A modern version of Project XL could invite companies to use AI in ways that go beyond minimum expectations. Instead of reacting after problems occur, companies could use predictive tools to prevent them. AI could help fine-tune emissions controls, better match clean energy supply with demand, reduce reliance on backup generators or detect leaks and equipment failures before they become serious. Pilot projects could also show how complex environmental and health data can be made clearer and more accessible to the public, helping communities better understand what’s happening around them.
For the EPA to truly play a constructive role in the AI era, it needs to do more than remove obstacles for industry and cut red tape.
Administrator Zeldin must empower the agency to use AI to defend the environment with the same zeal he applies to cutting the protections that safeguard it and our communities. A carefully designed, next-generation version of that earlier experiment could help the agency learn alongside the private sector, without losing sight of its public purpose.
Avi Garbow is the founder of consultancy Fiery Run Environmental Strategies and a nationally recognized environmental lawyer. He is the longest-serving general counsel at the U.S. Environmental Protection Agency, where he guided policy and strategy between 2009 and 2017.
Leave a Reply