Public Employees for Environmental Responsibility (PEER), Environment America, and DarkSky International want the Federal Communications Commission to require a broad, program wide environmental impact review of orbital data centers.
Three environmental groups asked the Federal Communications Commission this week to pause licensing of orbital data centers, betting that a 1970s environmental law most Americans associate with highways and pipelines can be turned against a planned million-satellite AI buildout in low Earth orbit.
Public Employees for Environmental Responsibility, Environment America, and DarkSky International filed a joint petition on or around July 8, 2026, asking the FCC to halt approvals of orbital data center projects, which the petition describes as AI compute hosted on large satellite constellations, until the agency completes a "programmatic" environmental impact statement under the National Environmental Policy Act. Earthjustice, the environmental law organization, is hosting the petition PDF and issuing the coalition press release as legal and communications counsel for the three groups.
NEPA is a procedural law: it does not ban anything. It requires federal agencies to take a "hard look" at the environmental and related social effects of "major federal actions" before greenlighting them, and to write up what they found. The petitioners are not arguing that orbital data centers are categorically illegal. They are arguing that the FCC cannot license them without first producing, and defending, a comprehensive analysis on the public record.
NEPA's power is in the pause, not the prohibition. A programmatic environmental impact statement can take years, and litigation over an inadequate one can take years more. For an industry that wants to start launching servers into orbit in the near term, that delay is the point.
The petitioners' case rests on two legal pillars. The first is NEPA itself: they argue that licensing a constellation of orbital data centers, as opposed to a single satellite, is a "major federal action" that triggers a higher-tier programmatic review rather than a series of narrow per-license evaluations. The second is the 1967 Outer Space Treaty's declaration that exploration and use of outer space is "the province of all mankind," which the petition invokes as grounds for additional scrutiny of any launch tied to orbital compute infrastructure.
What would actually defeat the petition is not a debate about ozone or night skies but a jurisdictional one. If the FCC concludes that satellite licensing is not a "major federal action," or that NEPA does not reach orbital data centers at all, the procedural brake disappears. Even an adverse ruling, on this theory, is a partial win: it produces an explicit FCC position the coalition can appeal or take to court.
The harms the coalition lists are concrete enough to give the legal question real weight. The petition flags ozone-layer degradation from launch and reentry emissions, altered stratospheric chemistry, the loss of night-sky quality that dark-sky advocates say is already accelerating with the current Starlink generation, and the cumulative impact of tens of thousands of reentering satellites at end of life. These are the kinds of effects NEPA was written to surface before a project is permitted, not after.
The scale is what makes the legal question urgent now. SpaceX has publicly described plans for up to a million Starlink-class satellites configured as orbital data centers, per the Gizmodo report and SpaceNews coverage. Blue Origin and the startup Starcloud are named in the petition as additional proponents. Independent reporting from The Register describes the broader orbital compute push as a gold rush now hitting an environmental review gate.
The coalition's central complaint is that the companies have refused to engage with the impact inquiry while promoting the technology as transformative. The petition and coalition materials argue that proponents have made sweeping public claims about the technology's scale and significance while declining to provide the data a programmatic review would require. That asymmetry, bold claims paired with no disclosure, is the gap the petition is designed to force closed.
The FCC's posture on the petition is not yet public. The agency now has an open docket, a defined set of petitioners with statutory standing under NEPA's "interested person" framework, and a statutory clock that does not actually run until the agency responds. The decision the agency makes over the next several months, whether orbital data center licensing is a "major federal action" requiring programmatic review or a routine series of per-license approvals, will determine whether the petitioners get a multi-year record to litigate or a procedural exit. The petitioners are not asking the FCC to decide whether orbital data centers should exist. They are asking the FCC to decide in writing, on a record, before the launch cadence catches up with the law.