The American Astronomical Society warned the FCC of flash blinded pilots and ruined observatories. The agency's order sets those concerns aside as outside its authority.
The FCC approved a satellite built to reflect concentrated sunlight onto nighttime ground targets, then used the same order to explain why assessing whether that light could blind pilots or ruin astronomical research is outside the agency's job.
The order, DA-26-706A1, grants spectrum authority to Reflect Orbital's Earendil-1, a single low-Earth-orbit demonstrator carrying an adjustable thin-film reflector and a propulsion module for collision avoidance and station-keeping. The license is radio-only. The agency's reasoning for setting the safety questions aside is now on the record: a regulator can greenlight a contested technology and immunize itself from the consequences by pointing to the exact edge of its jurisdiction. The FCC did that here.
Earendil-1 is a single demonstration satellite that has not yet launched. Reflect Orbital's longer commercial pitch, a Starlink-scale constellation selling on-demand nighttime sunlight for solar farms, search-and-rescue operations, and other applications, still belongs to a future filing, not this one. The FCC approved what was put in front of it and declined to adjudicate what was not.
The American Astronomical Society had told the FCC on May 29 that the demonstration, and any follow-on constellation, raised specific astronomy-related concerns. The filing cited interference with ground-based astronomy and "temporary flash blinding of pilots and drivers." Those are not stylistic complaints. They are public-safety and research-harms claims directed at a satellite that by design bounces concentrated light back at Earth.
The FCC's response, embedded in the same order, is that the Communications Act directs the agency to "encourage the provision of new technologies and services to the public," and that non-spectrum environmental and public-health effects lie outside the FCC's authority. The agency addressed the scientific criticism as part of its spectrum review and licensed the satellite anyway. Whether pilots can be blinded or observatories ruined by orbiting mirrors is, on this record, a question for someone else. The order does not name that someone else.
Reflect Orbital's own materials follow a similar line. The company, in its own FCC spectrum filing materials, markets nighttime sunlight as a service. It has also acknowledged, in those same materials, that eye damage is a risk for any telescope with an aperture larger than 12 inches. The threshold is a self-disclosed limit on the company's own product, not a regulatory finding. The fact that the company drew the line at all tells you where its own risk model ends and where the public-interest questions begin.
Wire coverage from Engadget and PCMag reads the news as a fight between the FCC and the astronomy community. That is the most direct rendering of the dispute. The order does something narrower: it resolved the radio question, recorded the harm warnings, and declined to weigh them on the merits. Knowing what was approved is the first half. Knowing why the FCC could grant the license without engaging the harm warnings is the other half.
The order is also the FCC's first formal spectrum decision on a sun-reflection demonstrator. LightNow reported Reflect Orbital's earlier spectrum petition in March. Engineers and regulators who track satellite light-pollution questions now have a public document that says, in effect, this decision set no precedent on safety. Each future application will have to make that argument again on its own record.
The next steps are specific. Earendil-1 has additional milestones to clear before launch. The order does not authorize any future constellation, and a separate application for a Starlink-scale reflector network would trigger a new proceeding, a new round of Ex Parte filings, and, depending on how the record develops, the same jurisdictional question on a larger scale. Whether any other agency steps into the gap the order left open is the unresolved piece of the public-interest ledger.