Ending the eternal night? The U.S. approves its first “space reflection mirror” satellite, which can reflect sunlight back to Earth

The U.S. FCC ignored more than 1,600 public comments and approved the world’s first commercial “space reflection mirror” satellite for launch, which can create “daylight” for specific regions.
(Background: SpaceX officially renamed itself SpaceXAI, and Musk tied space and AI together as one)
(Additional background: SpaceX destroyed 260 Starlink satellites in half a year! Burning up in the atmosphere has become routine, and environmental groups protest that it affects the ozone layer.)

Table of Contents

Toggle

• An 18-meter reflective mirror, calibrated every 4 minutes
• Even the FCC calls it an “odd space activity”
• Musk’s low Earth orbit is being turned into an AI gold-rush arena
• Flashes, biological clocks, and burning through the ozone layer
• Clean energy—or a “greenwashing” story?

After the U.S. Federal Communications Commission (FCC) ignored more than 1,600 public objections, it officially approved what is described as the first commercial “space reflection mirror” satellite in history. According to the FCC’s release documents, this satellite, code-named Earendil-1, is built by the startup Reflect Orbital. The mission is to reflect sunlight back to Earth, creating “daylight” for specific areas.

An 18-meter reflective mirror, calibrated every 4 minutes

Earendil-1 is a thin-film reflective mirror measuring 18 meters square, weighing 142 kilograms, and is expected to be launched later this year into an orbit 600 to 650 kilometers above the Earth. Its job is to reflect sunlight onto a surface area about 5 kilometers wide. Because the satellite continuously orbits the Earth, the reflection angle has to be recalibrated every 4 minutes to maintain the lighting effect.

This is only the first step in Reflect Orbital’s ambition. The company plans to deploy more than 50,000 similar satellites before 2035, claiming they can be applied in industrial scenarios such as agricultural supplemental lighting and emergency-response illumination. The American Astronomical Society (AAS) has raised opposition, arguing that reflected flashes could harm the eyes of amateur observers, and that the intense light could temporarily blind pilots and drivers in an instant.

Even the FCC calls it an “odd space activity”

The FCC recently released a document titled 《A Spectral Banquet of “Odd Space Activities”》—in plain terms, even the regulator itself is admitting that what is now being applied to go into orbit is so outlandish that it needs to be categorized separately.

The projects listed in the filing include: space advertising billboards, private space hotels for billionaires, artificial meteor-shower performances, a space burial service that sends ashes into orbit, and a series of orbital missile-program plans. At the same time, millions of orbital AI data center projects have also been put on the table. Companies are essentially competing to be the first to launch something “odd enough” to persuade investors to put up money.

Musk’s low Earth orbit is being turned into an AI gold-rush arena

At present, nearly 11,000 SpaceX Starlink satellites are already operating in orbit. Any company seeking to launch into low Earth orbit must first consider SpaceX’s deployments—sometimes even coordinating its flight path with SpaceX directly; otherwise there is a risk of collision. In December 2025, a Starlink satellite and a Chinese satellite nearly collided, prompting a scare. Even the launch schedules for Artemis I in 2022 and Artemis II in 2026 must deliberately avoid the Starlink satellite constellation, leaving only narrow launch windows.

Even more extraordinary: in February this year, SpaceX applied to the FCC to launch another 1,000,000 satellites, citing “use for AI data centers.” What does 1,000,000 satellites mean? It is equivalent to 40 times the total number of satellites humans have launched so far, and it is an entirely untested technology—indeed, it is uncertain whether it can operate normally in the space environment.

Not only did the FCC accept the application, the speed was astonishing. It gave scientists worldwide only 30 days to model and assess potential risks, even though key information—such as mass, size, materials, and orbital distribution—was extremely incomplete. As of now, at least 4 rival companies have followed suit. Each has applied for “AI data center” constellations on the scale of tens of thousands of satellites. SpaceX then added another application for 100,000 satellites to connect with its own 1,000,000-satellite data centers.

Flashes, biological clocks, and burning through the ozone layer

Behind these proposals lie concrete safety and environmental costs. Sudden intense flashes of light generated when calibrating mirror angles could temporarily deprive ground pilots and drivers of vision. Ongoing artificial light sources could also disrupt the biological clocks of humans, animals, and plants. Highly sensitive sensors on research telescopes, as well as star-tracker cameras on low Earth orbit satellites, could be damaged by reflected light that is too bright.

Going even further is a space solar power plan already locked in by Meta: using high-energy beams to send solar energy collected in space directly back to ground data centers. Such beams could change the composition of atmospheric chemistry, inadvertently harming birds and other wildlife that enter the beam path, and would also require establishing no-fly zones around receiving stations—while simultaneously avoiding the low Earth orbit flight paths of the Starlink satellites that SpaceX has just applied to launch.

And the metal particulates released when satellites burn up and re-enter the atmosphere at end of life have been preliminarily studied as something that would materially change atmospheric chemical components and erode the ozone layer. As more objects crowd into orbit, the risk of cascading collisions—known as the Kessler effect—also rises.

Clean energy—or a “greenwashing” story?

Almost all of these space projects market themselves under the banner of “clean energy.” But once you factor in the environmental costs of constructing, launching, maintaining, and ultimately burning up the satellites, the word “clean” doesn’t hold up—at its core, it is a greenwashing narrative. The FCC’s original mission is to manage the radio spectrum, yet it is now being forced to evaluate an issue of orbital safety that it may not have the professional judgment capability to assess. In theory, it should hand off part of the review to the U.S. Office of Space Commerce, but recent budget cuts make this option unrealistic.

Low Earth orbit may seem limitless, but it is actually extremely limited. Satellites circle the Earth every 90 minutes, and the likelihood of collisions is far higher than people imagine. The real question that should be asked is who should be held responsible for orbital congestion, atmospheric pollution, and even injuries on the ground. Doing more with fewer resources is the engineering challenge that this low Earth orbit AI gold rush truly should face—not who can send even more outrageous things into space first.