The short version

• On 9 October 2026 the FCC’s Space Bureau reissued Reflect Orbital’s licence for Earendil-1, a test satellite with a steerable mirror meant to reflect sunlight onto the ground at night. The change was narrow: the surety bond deadline moved to 5 January 2027 and a frequency typo was fixed.
• The July 2026 grant licensed the satellite’s radios and checked its debris plan, but the Bureau held that the reflector itself, and its effects on astronomy, wildlife and people, fall outside the FCC’s authority.
• An Application for Review filed by Earthjustice for DarkSky International and three other groups, which asks the full Commission to reverse the grant, was still pending when the Bureau acted on 9 October.
• The case exposes a familiar gap in US space law: a novel activity in orbit can be lawful to launch and lawful to talk to, yet no federal agency claims the job of reviewing what the payload actually does.

A company wants to put mirrors in orbit and sell sunlight after dark. The first of them, a single demonstration satellite called Earendil-1, now holds a US licence. On 9 October 2026 the Federal Communications Commission’s Space Bureau granted Reflect Orbital’s own petition for reconsideration and reissued that licence with a later bond deadline. It is a small order on its face, four pages and a reissued grant. What makes it worth reading is what it leaves untouched.

Read with the July grant it amends, the order confirms that the US government has authorised a spacecraft whose main function no agency claims the power to regulate. I think that matters more than the bond deadline.

What the FCC licensed in July

The original grant, DA 26-706, was adopted and released by the Space Bureau on 9 July 2026. It authorises Reflect Orbital to deploy and operate one satellite, call sign S00711, using UHF, S-band and X-band frequencies for telemetry, tracking, command and data downlinks. Earendil-1 is to be inserted at about 510 km and to operate at about 625 km in a near-polar orbit inclined at 88 degrees.

According to the Bureau’s summary of the application, the satellite carries a “deployable, highly specular thin-film reflector” that “is motorized and steerable to ensure reflected light is only visible in the targeted area.” The company describes Earendil-1 as a testbed for reflectors that could, among other things, “augment energy projects by extending usable hours for solar cells”.

The record against it was unusually large. The American Astronomical Society filed a petition to deny on 9 March 2026. The Bureau counted comments from many scientific and environmental organisations and letters from more than 1,800 individuals. It denied the AAS petition and granted the application with conditions. The licence term is two years, and the satellite must be launched and operating by 9 July 2032.

Why the Bureau said the mirror was not its business

The core of the July order is a jurisdictional holding. Section 301 of the Communications Act requires a licence to operate “any apparatus for the transmission of energy or communications or signals by radio”. The Bureau reasoned that its authorisation for Earendil-1 “is for the radio station”, together with the orbital debris plan that it reviews as part of the communications apparatus. Beyond those two things, it said, its licensing authority “does not inherently extend to all activities that a spacecraft conducts in space”.

It followed that reasoning through each objection. On optical astronomy, the Bureau said its rules “do not address the protection of optical astronomy” and declined the conditions the International Astronomical Union had asked for, including a ban on pointing the beam within 160 km of listed observatories. It treated them as redundant with the company’s commitment to coordinate with NASA, the National Science Foundation and astronomers. On eye safety, it said the issues were “far removed from the space station’s radiofrequency operations”. On glare, aviation and the prospect of tens of thousands of reflectors, it called the concerns hypothetical because only one satellite had been applied for. For completeness, it added that even if it could weigh these harms, they were unlikely to arise from a single test satellite.

The Bureau cited Loper Bright Enterprises v Raimondo (2024) and older cases holding that the FCC cannot use a general public interest standard to regulate beyond its statute, and said new obligations for a whole class of service belong in a rulemaking, not one-off licence conditions.

NEPA and the categorical exclusion, again

Environmental groups argued that the FCC had to prepare an Environmental Assessment under the National Environmental Policy Act. The Bureau disagreed on two levels. First, it held the grant categorically excluded under section 1.1306(a) of its rules, citing Seven County Infrastructure Coalition v Eagle County (2025) for the point that NEPA focuses on the “proposed action”, which here was approving a radio station. The reflector, it said, was “too attenuated” from that action and beyond its authority. Second, and in the alternative, it treated the comments as petitions for environmental review under section 1.1307(c) and found they did not show with specificity that a single test satellite may have a significant environmental effect.

This is the same structure the Commission used for SpaceX’s Gen2 constellation, and the D.C. Circuit approved a version of it in 2024, which I discussed in our note on Dark-Sky Association v FCC. The difference now is that the FCC is trying to take satellites out of NEPA altogether. Its August 2025 NEPA rulemaking notice (FCC 25-47) proposed that “space-based operations be excluded from NEPA because they are ‘extraterritorial activities’ with effects located entirely outside of the jurisdiction of the United States.” That proposal sits awkwardly with a satellite whose whole purpose is to put light on the ground inside the United States. If it is adopted as drafted, whether light reflected onto US soil is an effect “located entirely outside” the United States will probably be decided in litigation.

The petitions that followed

On 6 August 2026 Earthjustice filed an Application for Review on behalf of DarkSky International, Public Employees for Environmental Responsibility, Environment America and the American Bird Conservancy, asking the full Commission to overturn the Bureau. DarkSky’s announcement says the groups want a proper Environmental Assessment or Environmental Impact Statement and, failing that, 16 conditions set out in an annex. An individual filed a petition for reconsideration on 7 August. Reflect Orbital filed its own petition on 10 August.

On 27 August the groups asked Chairman Brendan Carr for a stay, saying they were “concerned that launch preparations may be imminent”.

What the 9 October order changed

The Order on Reconsideration, DA 26-1101, signed by Acting Space Bureau Chief Jennifer Gilsenan, deals only with Reflect Orbital’s petition. It does two things.

First, it moves the surety bond deadline. Under the Part 25 rules a licensee must post a bond within 30 days of grant, which put Reflect Orbital’s deadline at 10 August 2026. The company had asked in September 2025 to defer the bond, and the July grant never addressed that request. The Bureau waived the August deadline on its own motion and set a new one of 5 January 2027, 180 days from grant, in line with recent practice for small systems. It said Reflect Orbital “is scheduled to launch the single Earendil-1 satellite within several months of grant”, which removes the warehousing concern the bond exists to address. It refused the company’s alternative request for a permanent deferral, pointing to the new Part 100 rules adopted in the Space Modernization proceeding and the Bureau’s 7 October transition guidance, which we covered in our piece on FCC Part 100 milestones and surety bonds.

Second, it corrects a typographical error in the centre frequency listed in Condition 8A, which governs uplinks in the 2025 to 2110 MHz band. Apart from those two conditions, the reissued grant matches the July version.

Just as important is what the order does not do. A footnote states that the individual petition and the Earthjustice Application for Review “address different issues” and that “both remain pending”. Nothing in the 9 October order touches the jurisdiction or NEPA questions.

The gap the order leaves open

The international obligation here is clear. Article VI of the Outer Space Treaty requires that the activities of non-governmental entities in outer space “require authorization and continuing supervision by the appropriate State Party”. Article IX adds a duty of due regard and of consultation where an activity could cause harmful interference with the activities of other States. Neither article is limited to radio.

In the United States, the FAA licenses the launch, the FCC licenses the radios and NOAA licenses remote sensing cameras. For anything else the satellite does, there is no general mission authorisation regime, a problem I looked at in the context of in-orbit servicing in our piece on the Swift rescue mission. Earendil-1 is a cleaner example than most, because its payload is not a sensor or a robotic arm but a light source aimed at the Earth.

There is one checkpoint that may still matter. If Earendil-1 flies on a launch licensed by the FAA, the payload is subject to the FAA’s review under 14 CFR 450.43, which asks whether its launch “would not jeopardize public health and safety, safety of property, U.S. national security or foreign policy interests, or international obligations of the United States.” The rule excludes aspects of a payload “subject to regulation by the Federal Communications Commission”. The FCC has now said the reflector is not subject to its regulation, so on the FCC’s own reasoning it falls into the space the FAA rule leaves open. I have seen no public sign that the FAA intends to use payload review this way, and it was not designed as an environmental or astronomy review.

What I would watch

The first is the Application for Review. The full Commission can affirm the Bureau, reverse it or send it back. Given the Commission’s NEPA proposal, I would be surprised by a reversal, but a Commission-level order would give the groups a final decision they can take to the D.C. Circuit, and the 2024 Dark-Sky case shows that these groups are prepared to litigate.

The second is the bond date and the launch. A bond posted by 5 January 2027 and a launch soon after would make Earendil-1 a test of the technology before any court has ruled on the licence. Once the satellite is in orbit, a challenge becomes much harder to turn into practical relief.

The third is the second application. The Bureau was explicit that its grant “does not predetermine action on any future application”, and Reflect Orbital has itself acknowledged that a constellation may need more review. When a multi-satellite application arrives, the “hypothetical” label the Bureau used in July will no longer be available, and the question of who regulates the light will be unavoidable.

The fourth is abroad. The IAU and foreign astronomical societies objected. If a beam is ever pointed at, or scatters over, another country, Articles VI and IX raise questions a radio licence does not answer.