The Cheapest Moat Is the Rocket. The Impossible One Is the Paperwork.

The U.S. Federal Communications Commission is Starlink's primary licensing authority, and the grants came in stages. The original Gen1 constellation — 4,425 satellites in Ku- and Ka-band — was authorized in March 2018. In December 2022 the FCC approved the Gen2 system with an initial 7,500 satellites, a partial grant of the roughly 29,988 SpaceX had requested, choosing to phase deployment rather than approve the full number at once. Trade-press reporting indicates a January 2026 expansion that added on the order of another 7,500 satellites toward 15,000 Gen2, along with additional gateway bands and lower-altitude orbital shells; we treat that most recent step as reported rather than confirmed, because outlets disagree on the precise date and band specifics and we prefer to verify against the primary FCC order before stating it as fact. The throughline is that each authorization is a regulator granting exclusive operating rights over a slice of a public resource — and the cumulative grant, built over years, is one of the assets a competitor would have to replicate from zero. Internationally, the ITU coordinates the underlying frequency and orbital filings, where SpaceX's early priority dates matter enormously.

Starlink operates across several bands, each serving a distinct function in the network. The user links — the connection between a customer's dish and the satellites — run in Ku-band and Ka-band, the established workhorse frequencies for satellite broadband, balancing capacity against rain-fade resilience. The gateway links, which connect satellites to ground stations and carry aggregated traffic, use higher-frequency V-band and E-band (around 71–86 GHz), where there is more bandwidth available for the heavy backhaul a large constellation needs. Between satellites, Starlink increasingly relies on laser inter-satellite links, which carry traffic optically and avoid consuming radio spectrum altogether — an important detail, because it lets the constellation route globally without coordinating ground spectrum for every hop. For direct-to-cell service, Starlink uses terrestrial mobile spectrum leased from a partner carrier rather than its own satellite bands. The strategic point is that holding licensed rights across this stack — consumer Ku/Ka, gateway V/E, and a direct-to-cell carve-out — is a multi-year regulatory achievement, not a purchase. A rival needs not one frequency but a coordinated portfolio of them, in the right orbits, secured before someone else claims them.

Through a regulatory mechanism called Supplemental Coverage from Space (SCS), which lets a satellite transmit signals that an unmodified, standard mobile phone can receive as if they came from a terrestrial cell tower. Starlink does this by leasing mobile spectrum from a carrier partner — in the U.S., T-Mobile's PCS G-block at 1910–1915 and 1990–1995 MHz — so the phone simply sees a familiar mobile frequency and connects without any special hardware or antenna. The FCC granted the first-ever SCS authorization for this arrangement in November 2024 and subsequently raised the permitted power levels in March 2025, both regulatory firsts that effectively created the rules for satellite-to-phone service as Starlink deployed it. That sequence is itself the moat: a competitor cannot simply replicate direct-to-cell, because doing so requires three things at once — a low-orbit satellite constellation capable of the link, a national mobile carrier willing to share its licensed terrestrial spectrum, and a novel regulatory approval that did not exist until Starlink and the FCC built it together. The technology is reproducible; the spectrum-and-partnership package is not, at least not quickly.

Not in any straightforward way, and that is the core of the moat. Spectrum and orbital slots are not freely traded commodities; they are coordinated, finite public resources allocated through national regulators (like the FCC) and the International Telecommunication Union, which operates largely on a first-come, first-served basis — the date a filing's advance-publication information is received sets its precedence. SpaceX filed early and deployed aggressively, so it holds priority in the most useful orbital shells and frequency coordinations. A well-funded rival such as Amazon's Kuiper can obtain its own licenses, and Kuiper has, but it inherits a later priority date and must coordinate around the incumbent already occupying the shells and bands it wants. It cannot buy SpaceX's priority, and it cannot compress the multi-year international coordination process with capital. This is why the paperwork is a harder moat than the rocket: launch capacity can be built, but a 2018-era orbital priority date cannot be purchased in 2026. The two operators have also filed mutual orbital-debris and interference complaints at the FCC — regulatory friction that, in practice, functions as a competitive tool to slow each other's deployment.

The 12 GHz fight was a multi-year regulatory battle over whether the 12 GHz band should be opened to high-power terrestrial 5G use. EchoStar/Dish and RS Access argued for repurposing the band for ground-based mobile broadband, while SpaceX argued that doing so would create interference for Starlink's satellite downlinks, which use the band to deliver service to customers. The stakes were real: a decision to flood 12 GHz with terrestrial transmitters could have degraded the quality of Starlink's consumer links across the United States. The FCC's 2023 resolution largely protected the satellite operators: it preserved 12.2 GHz for non-geostationary satellite use while freeing the adjacent 12.7 GHz for terrestrial services, a split that let both sides claim partial victory but kept Starlink's core downlink spectrum intact. Reporting since has indicated further spectrum maneuvering, including a move by EchoStar to transfer holdings toward terrestrial partners; we treat the specifics as reported pending confirmation. The episode matters because it shows the moat is defended, not just held — Starlink had to fight, at the FCC, to keep the spectrum its consumer service depends on, and it largely won.