Tech Companies Face Accusations of Avoiding Pollution Rules for AI Datacenters

The rapid expansion of artificial intelligence (AI) is creating a new conflict between big tech companies and the communities where they build their massive datacenters. Following recent reports of global protests against these energy-hungry facilities, new evidence suggests that developers are using questionable methods to avoid environmental regulations in the United States. Environmental advocates now accuse these companies of manipulating the permitting process to bypass strict air pollution rules.

Under the Clean Air Act, projects that are expected to be "major" sources of pollution must undergo a rigorous review process. This process is designed to protect local air quality by ensuring that new facilities do not violate safety limits. However, according to The Guardian, some developers are avoiding these major reviews by dividing their projects into multiple smaller, "minor" sources. By submitting several separate permit applications instead of one large one, companies can often avoid public scrutiny and get their projects approved much faster.

Michael Koerber, a retired official from the Environmental Protection Agency (EPA), described this practice as "lego permitting." He explained that by breaking a large project into smaller pieces, companies can effectively hide the true amount of pollution the facility will create. Koerber, who now works with the Environmental Protection Network, noted that these companies often hire expert consultants who know exactly how to find these regulatory shortcuts. "They want to fly under the radar," he said.

Several examples of this practice have surfaced across the country. In North Carolina, Amazon and Duke Energy submitted two separate permits for a project involving 649 diesel generators. By doing so, they avoided a major review, even though the facility is located in an area that already suffers from poor air quality. Similarly, in Maryland, multiple datacenters—including one by Amazon—are operating on a single campus under separate minor permits. In New Mexico, developers of the Project Jupiter datacenter also attempted to use this strategy by proposing two separate natural gas power sources to avoid a major environmental assessment.

In response to these allegations, an Amazon spokesperson stated that the company "appropriately applied for and obtained an air permit" for its projects. However, the use of such tactics has fueled growing public anger. This frustration is part of a larger, nationwide backlash against the AI industry. As reported by The Guardian, many Americans are increasingly concerned about the costs associated with these massive computer centers, which require vast amounts of electricity and water to function.

Public opinion data reflects this shift in attitude. A recent survey by the Marquette University Law School found that 73% of Americans believe the costs of datacenters outweigh their benefits. Furthermore, 70% of those surveyed expressed the view that AI is bad for society. This is a significant change from just a few years ago, when public excitement about new technology was much higher. Today, even younger adults, who are typically the most supportive of new tech, are becoming skeptical.

This widespread distrust is not just about local environmental concerns. Many people feel that a small group of powerful tech leaders is pushing a "civilization-threatening" project forward without public consent. There is a deep, growing suspicion of a political system that allows these companies to operate behind the scenes. As communities across the country continue to organize, they are increasingly calling for moratoria—a temporary ban on activity—and stricter rules to stop the unchecked growth of these facilities.

As the industry moves forward, the pressure on tech companies to be more transparent is likely to increase. The ability to "fly under the radar" is becoming more difficult as residents, local governments, and environmental groups demand more accountability. Whether these companies can continue to use these regulatory shortcuts remains to be seen, but the era of building AI infrastructure without significant public pushback appears to be over. For now, the conflict between the rapid development of AI and the needs of local communities remains one of the most important issues in the American energy landscape.